{"id": "wf12:laboratory_batch:test:0000:choice", "group_id": "wf12:laboratory_batch:test:0000", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fc2d9c607b.\\nD means contamination >= 70.\\nB means traceability >= 87.\\nC means supervision >= 76.\\nA means purity >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 48, \\\"traceability\\\": 29, \\\"supervision\\\": 85, \\\"contamination\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2345e0\": \"Outcome PROCEED.\", \"candidate_3f1bf9\": \"Outcome REVIEW.\", \"candidate_53e896\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2345e0\": 1.0, \"candidate_3f1bf9\": 0.0, \"candidate_53e896\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0000:score", "group_id": "wf12:laboratory_batch:test:0000", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fc2d9c607b.\\nD means contamination >= 70.\\nB means traceability >= 87.\\nC means supervision >= 76.\\nA means purity >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 48, \\\"traceability\\\": 29, \\\"supervision\\\": 85, \\\"contamination\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0000:noul", "group_id": "wf12:laboratory_batch:test:0000", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fc2d9c607b.\\nD means contamination >= 70.\\nB means traceability >= 87.\\nC means supervision >= 76.\\nA means purity >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 48, \\\"traceability\\\": 29, \\\"supervision\\\": 85, \\\"contamination\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0000:uncertain", "group_id": "wf12:laboratory_batch:test:0000", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fc2d9c607b.\\nD means contamination >= 70.\\nB means traceability >= 87.\\nC means supervision >= 76.\\nA means purity >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 48, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 85, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2345e0\": \"Outcome PROCEED.\", \"candidate_3f1bf9\": \"Outcome REVIEW.\", \"candidate_53e896\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2345e0\": 0.65, \"candidate_3f1bf9\": 0.35, \"candidate_53e896\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0001:choice", "group_id": "wf12:laboratory_batch:test:0001", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 129d514aab.\\nC means supervision >= 71.\\nA means purity >= 20.\\nD means contamination >= 36.\\nB means traceability >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 96, \\\"supervision\\\": 66, \\\"contamination\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e41829\": \"Outcome PROCEED.\", \"candidate_2eaa70\": \"Outcome REVIEW.\", \"candidate_0eac4b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e41829\": 0.0, \"candidate_2eaa70\": 1.0, \"candidate_0eac4b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0001:score", "group_id": "wf12:laboratory_batch:test:0001", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 129d514aab.\\nC means supervision >= 71.\\nA means purity >= 20.\\nD means contamination >= 36.\\nB means traceability >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 96, \\\"supervision\\\": 66, \\\"contamination\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0001:noul", "group_id": "wf12:laboratory_batch:test:0001", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 129d514aab.\\nC means supervision >= 71.\\nA means purity >= 20.\\nD means contamination >= 36.\\nB means traceability >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 96, \\\"supervision\\\": 66, \\\"contamination\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0001:uncertain", "group_id": "wf12:laboratory_batch:test:0001", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 129d514aab.\\nC means supervision >= 71.\\nA means purity >= 20.\\nD means contamination >= 36.\\nB means traceability >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 96, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 0}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e41829\": \"Outcome PROCEED.\", \"candidate_2eaa70\": \"Outcome REVIEW.\", \"candidate_0eac4b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e41829\": 0.9073972602739726, \"candidate_2eaa70\": 0.0926027397260274, \"candidate_0eac4b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0002:choice", "group_id": "wf12:laboratory_batch:test:0002", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b7ab7f8c9e.\\nA means purity >= 55.\\nC means supervision >= 88.\\nB means traceability >= 58.\\nD means contamination >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 71, \\\"supervision\\\": 65, \\\"contamination\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ea9653\": \"Outcome PROCEED.\", \"candidate_195011\": \"Outcome REVIEW.\", \"candidate_6d1481\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ea9653\": 0.0, \"candidate_195011\": 0.0, \"candidate_6d1481\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0002:score", "group_id": "wf12:laboratory_batch:test:0002", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b7ab7f8c9e.\\nA means purity >= 55.\\nC means supervision >= 88.\\nB means traceability >= 58.\\nD means contamination >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 71, \\\"supervision\\\": 65, \\\"contamination\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0002:noul", "group_id": "wf12:laboratory_batch:test:0002", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b7ab7f8c9e.\\nA means purity >= 55.\\nC means supervision >= 88.\\nB means traceability >= 58.\\nD means contamination >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 71, \\\"supervision\\\": 65, \\\"contamination\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0002:uncertain", "group_id": "wf12:laboratory_batch:test:0002", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b7ab7f8c9e.\\nA means purity >= 55.\\nC means supervision >= 88.\\nB means traceability >= 58.\\nD means contamination >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 74}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ea9653\": \"Outcome PROCEED.\", \"candidate_195011\": \"Outcome REVIEW.\", \"candidate_6d1481\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ea9653\": 0.0, \"candidate_195011\": 0.1, \"candidate_6d1481\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0003:choice", "group_id": "wf12:laboratory_batch:test:0003", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f85fafccc8.\\nB means traceability >= 22.\\nA means purity >= 82.\\nC means supervision >= 73.\\nD means contamination >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 90, \\\"supervision\\\": 76, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bcb17a\": \"Outcome PROCEED.\", \"candidate_98ad04\": \"Outcome REVIEW.\", \"candidate_71522a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bcb17a\": 1.0, \"candidate_98ad04\": 0.0, \"candidate_71522a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0003:score", "group_id": "wf12:laboratory_batch:test:0003", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f85fafccc8.\\nB means traceability >= 22.\\nA means purity >= 82.\\nC means supervision >= 73.\\nD means contamination >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 90, \\\"supervision\\\": 76, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0003:noul", "group_id": "wf12:laboratory_batch:test:0003", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f85fafccc8.\\nB means traceability >= 22.\\nA means purity >= 82.\\nC means supervision >= 73.\\nD means contamination >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 90, \\\"supervision\\\": 76, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0003:uncertain", "group_id": "wf12:laboratory_batch:test:0003", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f85fafccc8.\\nB means traceability >= 22.\\nA means purity >= 82.\\nC means supervision >= 73.\\nD means contamination >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 76, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.65, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0004:choice", "group_id": "wf12:laboratory_batch:test:0004", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9d130858bc.\\nD means contamination >= 52.\\nA means purity >= 38.\\nC means supervision >= 51.\\nB means traceability >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 74, \\\"traceability\\\": 62, \\\"supervision\\\": 79, \\\"contamination\\\": 59}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a95a1d\": \"Outcome PROCEED.\", \"candidate_7723b9\": \"Outcome REVIEW.\", \"candidate_e6465b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a95a1d\": 0.0, \"candidate_7723b9\": 1.0, \"candidate_e6465b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0004:score", "group_id": "wf12:laboratory_batch:test:0004", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9d130858bc.\\nD means contamination >= 52.\\nA means purity >= 38.\\nC means supervision >= 51.\\nB means traceability >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 74, \\\"traceability\\\": 62, \\\"supervision\\\": 79, \\\"contamination\\\": 59}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0004:noul", "group_id": "wf12:laboratory_batch:test:0004", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9d130858bc.\\nD means contamination >= 52.\\nA means purity >= 38.\\nC means supervision >= 51.\\nB means traceability >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 74, \\\"traceability\\\": 62, \\\"supervision\\\": 79, \\\"contamination\\\": 59}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0004:uncertain", "group_id": "wf12:laboratory_batch:test:0004", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9d130858bc.\\nD means contamination >= 52.\\nA means purity >= 38.\\nC means supervision >= 51.\\nB means traceability >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 74, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 79, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.65, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0005:choice", "group_id": "wf12:laboratory_batch:test:0005", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2cc41acd09.\\nA means purity >= 43.\\nB means traceability >= 33.\\nD means contamination >= 28.\\nC means supervision >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 61, \\\"traceability\\\": 75, \\\"supervision\\\": 49, \\\"contamination\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_85a7f9\": \"Outcome PROCEED.\", \"candidate_47c353\": \"Outcome REVIEW.\", \"candidate_65f675\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_85a7f9\": 0.0, \"candidate_47c353\": 0.0, \"candidate_65f675\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0005:score", "group_id": "wf12:laboratory_batch:test:0005", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2cc41acd09.\\nA means purity >= 43.\\nB means traceability >= 33.\\nD means contamination >= 28.\\nC means supervision >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 61, \\\"traceability\\\": 75, \\\"supervision\\\": 49, \\\"contamination\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0005:noul", "group_id": "wf12:laboratory_batch:test:0005", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2cc41acd09.\\nA means purity >= 43.\\nB means traceability >= 33.\\nD means contamination >= 28.\\nC means supervision >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 61, \\\"traceability\\\": 75, \\\"supervision\\\": 49, \\\"contamination\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0005:uncertain", "group_id": "wf12:laboratory_batch:test:0005", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2cc41acd09.\\nA means purity >= 43.\\nB means traceability >= 33.\\nD means contamination >= 28.\\nC means supervision >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 61, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 37}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.1, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0006:choice", "group_id": "wf12:laboratory_batch:test:0006", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c1a46a8176.\\nB means traceability >= 56.\\nD means contamination >= 82.\\nA means purity >= 37.\\nC means supervision >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 72, \\\"supervision\\\": 80, \\\"contamination\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e2ecf5\": \"Outcome PROCEED.\", \"candidate_2f1fbb\": \"Outcome REVIEW.\", \"candidate_732ea8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e2ecf5\": 1.0, \"candidate_2f1fbb\": 0.0, \"candidate_732ea8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0006:score", "group_id": "wf12:laboratory_batch:test:0006", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c1a46a8176.\\nB means traceability >= 56.\\nD means contamination >= 82.\\nA means purity >= 37.\\nC means supervision >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 72, \\\"supervision\\\": 80, \\\"contamination\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0006:noul", "group_id": "wf12:laboratory_batch:test:0006", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c1a46a8176.\\nB means traceability >= 56.\\nD means contamination >= 82.\\nA means purity >= 37.\\nC means supervision >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 72, \\\"supervision\\\": 80, \\\"contamination\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0006:uncertain", "group_id": "wf12:laboratory_batch:test:0006", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c1a46a8176.\\nB means traceability >= 56.\\nD means contamination >= 82.\\nA means purity >= 37.\\nC means supervision >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.2, \\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.93, \"true\": 0.06999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0007:choice", "group_id": "wf12:laboratory_batch:test:0007", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fbc22c144c.\\nA means purity >= 39.\\nB means traceability >= 29.\\nD means contamination >= 52.\\nC means supervision >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": 82, \\\"supervision\\\": 84, \\\"contamination\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_24d063\": \"Outcome PROCEED.\", \"candidate_c84d4d\": \"Outcome REVIEW.\", \"candidate_a064cb\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_24d063\": 0.0, \"candidate_c84d4d\": 1.0, \"candidate_a064cb\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0007:score", "group_id": "wf12:laboratory_batch:test:0007", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fbc22c144c.\\nA means purity >= 39.\\nB means traceability >= 29.\\nD means contamination >= 52.\\nC means supervision >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": 82, \\\"supervision\\\": 84, \\\"contamination\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0007:noul", "group_id": "wf12:laboratory_batch:test:0007", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fbc22c144c.\\nA means purity >= 39.\\nB means traceability >= 29.\\nD means contamination >= 52.\\nC means supervision >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": 82, \\\"supervision\\\": 84, \\\"contamination\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0007:uncertain", "group_id": "wf12:laboratory_batch:test:0007", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fbc22c144c.\\nA means purity >= 39.\\nB means traceability >= 29.\\nD means contamination >= 52.\\nC means supervision >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.35, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.13}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0008:choice", "group_id": "wf12:laboratory_batch:test:0008", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 055865cbae.\\nC means supervision >= 13.\\nA means purity >= 23.\\nD means contamination >= 56.\\nB means traceability >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": 26, \\\"supervision\\\": 7, \\\"contamination\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_812c27\": \"Outcome PROCEED.\", \"candidate_a71740\": \"Outcome REVIEW.\", \"candidate_2cfc9a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_812c27\": 0.0, \"candidate_a71740\": 0.0, \"candidate_2cfc9a\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0008:score", "group_id": "wf12:laboratory_batch:test:0008", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 055865cbae.\\nC means supervision >= 13.\\nA means purity >= 23.\\nD means contamination >= 56.\\nB means traceability >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": 26, \\\"supervision\\\": 7, \\\"contamination\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0008:noul", "group_id": "wf12:laboratory_batch:test:0008", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 055865cbae.\\nC means supervision >= 13.\\nA means purity >= 23.\\nD means contamination >= 56.\\nB means traceability >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": 26, \\\"supervision\\\": 7, \\\"contamination\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0008:uncertain", "group_id": "wf12:laboratory_batch:test:0008", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 055865cbae.\\nC means supervision >= 13.\\nA means purity >= 23.\\nD means contamination >= 56.\\nB means traceability >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 26, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 79}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.95, \"true\": 0.050000000000000044}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0009:choice", "group_id": "wf12:laboratory_batch:test:0009", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5ddf253002.\\nC means supervision >= 35.\\nD means contamination >= 17.\\nA means purity >= 79.\\nB means traceability >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 96, \\\"traceability\\\": 8, \\\"supervision\\\": 57, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_434078\": \"Outcome PROCEED.\", \"candidate_20b302\": \"Outcome REVIEW.\", \"candidate_00d93e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_434078\": 1.0, \"candidate_20b302\": 0.0, \"candidate_00d93e\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0009:score", "group_id": "wf12:laboratory_batch:test:0009", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5ddf253002.\\nC means supervision >= 35.\\nD means contamination >= 17.\\nA means purity >= 79.\\nB means traceability >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 96, \\\"traceability\\\": 8, \\\"supervision\\\": 57, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0009:noul", "group_id": "wf12:laboratory_batch:test:0009", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5ddf253002.\\nC means supervision >= 35.\\nD means contamination >= 17.\\nA means purity >= 79.\\nB means traceability >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 96, \\\"traceability\\\": 8, \\\"supervision\\\": 57, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0009:uncertain", "group_id": "wf12:laboratory_batch:test:0009", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5ddf253002.\\nC means supervision >= 35.\\nD means contamination >= 17.\\nA means purity >= 79.\\nB means traceability >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.2, \\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_434078\": \"Outcome PROCEED.\", \"candidate_20b302\": \"Outcome REVIEW.\", \"candidate_00d93e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_434078\": 0.4589000000000001, \"candidate_20b302\": 0.4186, \"candidate_00d93e\": 0.12250000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0010:choice", "group_id": "wf12:laboratory_batch:test:0010", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 57aaa82d78.\\nC means supervision >= 57.\\nB means traceability >= 64.\\nA means purity >= 35.\\nD means contamination >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 72, \\\"supervision\\\": 78, \\\"contamination\\\": 59}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f3b8eb\": \"Outcome PROCEED.\", \"candidate_bf1c55\": \"Outcome REVIEW.\", \"candidate_60eef8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f3b8eb\": 0.0, \"candidate_bf1c55\": 1.0, \"candidate_60eef8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0010:score", "group_id": "wf12:laboratory_batch:test:0010", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 57aaa82d78.\\nC means supervision >= 57.\\nB means traceability >= 64.\\nA means purity >= 35.\\nD means contamination >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 72, \\\"supervision\\\": 78, \\\"contamination\\\": 59}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0010:noul", "group_id": "wf12:laboratory_batch:test:0010", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 57aaa82d78.\\nC means supervision >= 57.\\nB means traceability >= 64.\\nA means purity >= 35.\\nD means contamination >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 72, \\\"supervision\\\": 78, \\\"contamination\\\": 59}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0010:uncertain", "group_id": "wf12:laboratory_batch:test:0010", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 57aaa82d78.\\nC means supervision >= 57.\\nB means traceability >= 64.\\nA means purity >= 35.\\nD means contamination >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 72, \\\"supervision\\\": 78, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f3b8eb\": \"Outcome PROCEED.\", \"candidate_bf1c55\": \"Outcome REVIEW.\", \"candidate_60eef8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f3b8eb\": 0.65, \"candidate_bf1c55\": 0.35, \"candidate_60eef8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0011:choice", "group_id": "wf12:laboratory_batch:test:0011", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dfa912f739.\\nA means purity >= 13.\\nC means supervision >= 62.\\nB means traceability >= 32.\\nD means contamination >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 0, \\\"traceability\\\": 17, \\\"supervision\\\": 21, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d04bb6\": \"Outcome PROCEED.\", \"candidate_4310c9\": \"Outcome REVIEW.\", \"candidate_be5873\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d04bb6\": 0.0, \"candidate_4310c9\": 0.0, \"candidate_be5873\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0011:score", "group_id": "wf12:laboratory_batch:test:0011", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dfa912f739.\\nA means purity >= 13.\\nC means supervision >= 62.\\nB means traceability >= 32.\\nD means contamination >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 0, \\\"traceability\\\": 17, \\\"supervision\\\": 21, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0011:noul", "group_id": "wf12:laboratory_batch:test:0011", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dfa912f739.\\nA means purity >= 13.\\nC means supervision >= 62.\\nB means traceability >= 32.\\nD means contamination >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 0, \\\"traceability\\\": 17, \\\"supervision\\\": 21, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0011:uncertain", "group_id": "wf12:laboratory_batch:test:0011", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dfa912f739.\\nA means purity >= 13.\\nC means supervision >= 62.\\nB means traceability >= 32.\\nD means contamination >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 0, \\\"traceability\\\": 17, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.1, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d04bb6\": \"Outcome PROCEED.\", \"candidate_4310c9\": \"Outcome REVIEW.\", \"candidate_be5873\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d04bb6\": 0.009999999999999998, \"candidate_4310c9\": 0.18, \"candidate_be5873\": 0.81}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0012:choice", "group_id": "wf12:laboratory_batch:test:0012", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3a3d521fad.\\nC means supervision >= 78.\\nA means purity >= 61.\\nD means contamination >= 75.\\nB means traceability >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 100, \\\"supervision\\\": 82, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_90790b\": \"Outcome PROCEED.\", \"candidate_b5d606\": \"Outcome REVIEW.\", \"candidate_1802fc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_90790b\": 1.0, \"candidate_b5d606\": 0.0, \"candidate_1802fc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0012:score", "group_id": "wf12:laboratory_batch:test:0012", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3a3d521fad.\\nC means supervision >= 78.\\nA means purity >= 61.\\nD means contamination >= 75.\\nB means traceability >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 100, \\\"supervision\\\": 82, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0012:noul", "group_id": "wf12:laboratory_batch:test:0012", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3a3d521fad.\\nC means supervision >= 78.\\nA means purity >= 61.\\nD means contamination >= 75.\\nB means traceability >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 100, \\\"supervision\\\": 82, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0012:uncertain", "group_id": "wf12:laboratory_batch:test:0012", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3a3d521fad.\\nC means supervision >= 78.\\nA means purity >= 61.\\nD means contamination >= 75.\\nB means traceability >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 67}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9600000000000001, \"1\": 0.03999999999999998, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0013:choice", "group_id": "wf12:laboratory_batch:test:0013", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 65e600d309.\\nA means purity >= 17.\\nB means traceability >= 40.\\nD means contamination >= 55.\\nC means supervision >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 15, \\\"traceability\\\": 4, \\\"supervision\\\": 56, \\\"contamination\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_98048a\": \"Outcome PROCEED.\", \"candidate_122f5d\": \"Outcome REVIEW.\", \"candidate_7a6234\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_98048a\": 0.0, \"candidate_122f5d\": 1.0, \"candidate_7a6234\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0013:score", "group_id": "wf12:laboratory_batch:test:0013", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 65e600d309.\\nA means purity >= 17.\\nB means traceability >= 40.\\nD means contamination >= 55.\\nC means supervision >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 15, \\\"traceability\\\": 4, \\\"supervision\\\": 56, \\\"contamination\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0013:noul", "group_id": "wf12:laboratory_batch:test:0013", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 65e600d309.\\nA means purity >= 17.\\nB means traceability >= 40.\\nD means contamination >= 55.\\nC means supervision >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 15, \\\"traceability\\\": 4, \\\"supervision\\\": 56, \\\"contamination\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0013:uncertain", "group_id": "wf12:laboratory_batch:test:0013", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 65e600d309.\\nA means purity >= 17.\\nB means traceability >= 40.\\nD means contamination >= 55.\\nC means supervision >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 4, \\\"supervision\\\": 56, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.65, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0014:choice", "group_id": "wf12:laboratory_batch:test:0014", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9cf6752caf.\\nB means traceability >= 38.\\nD means contamination >= 87.\\nA means purity >= 37.\\nC means supervision >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 77, \\\"supervision\\\": 4, \\\"contamination\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_57fa9a\": \"Outcome PROCEED.\", \"candidate_0599fb\": \"Outcome REVIEW.\", \"candidate_73ff3c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_57fa9a\": 0.0, \"candidate_0599fb\": 0.0, \"candidate_73ff3c\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0014:score", "group_id": "wf12:laboratory_batch:test:0014", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9cf6752caf.\\nB means traceability >= 38.\\nD means contamination >= 87.\\nA means purity >= 37.\\nC means supervision >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 77, \\\"supervision\\\": 4, \\\"contamination\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0014:noul", "group_id": "wf12:laboratory_batch:test:0014", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9cf6752caf.\\nB means traceability >= 38.\\nD means contamination >= 87.\\nA means purity >= 37.\\nC means supervision >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 77, \\\"supervision\\\": 4, \\\"contamination\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0014:uncertain", "group_id": "wf12:laboratory_batch:test:0014", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9cf6752caf.\\nB means traceability >= 38.\\nD means contamination >= 87.\\nA means purity >= 37.\\nC means supervision >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 90}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"C\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.9, \"2\": 0.09999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0015:choice", "group_id": "wf12:laboratory_batch:test:0015", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9b7bae97fe.\\nC means supervision >= 48.\\nB means traceability >= 60.\\nD means contamination >= 84.\\nA means purity >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 53, \\\"supervision\\\": 98, \\\"contamination\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9efcdd\": \"Outcome PROCEED.\", \"candidate_efbf28\": \"Outcome REVIEW.\", \"candidate_6e39c2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9efcdd\": 1.0, \"candidate_efbf28\": 0.0, \"candidate_6e39c2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0015:score", "group_id": "wf12:laboratory_batch:test:0015", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9b7bae97fe.\\nC means supervision >= 48.\\nB means traceability >= 60.\\nD means contamination >= 84.\\nA means purity >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 53, \\\"supervision\\\": 98, \\\"contamination\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0015:noul", "group_id": "wf12:laboratory_batch:test:0015", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9b7bae97fe.\\nC means supervision >= 48.\\nB means traceability >= 60.\\nD means contamination >= 84.\\nA means purity >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 53, \\\"supervision\\\": 98, \\\"contamination\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0015:uncertain", "group_id": "wf12:laboratory_batch:test:0015", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9b7bae97fe.\\nC means supervision >= 48.\\nB means traceability >= 60.\\nD means contamination >= 84.\\nA means purity >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 53, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.3599999999999999, \"true\": 0.6400000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0016:choice", "group_id": "wf12:laboratory_batch:test:0016", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ac0f4e5388.\\nA means purity >= 70.\\nD means contamination >= 75.\\nC means supervision >= 29.\\nB means traceability >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 97, \\\"supervision\\\": 54, \\\"contamination\\\": 77}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eed465\": \"Outcome PROCEED.\", \"candidate_a1d9f3\": \"Outcome REVIEW.\", \"candidate_52e421\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eed465\": 0.0, \"candidate_a1d9f3\": 1.0, \"candidate_52e421\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0016:score", "group_id": "wf12:laboratory_batch:test:0016", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ac0f4e5388.\\nA means purity >= 70.\\nD means contamination >= 75.\\nC means supervision >= 29.\\nB means traceability >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 97, \\\"supervision\\\": 54, \\\"contamination\\\": 77}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0016:noul", "group_id": "wf12:laboratory_batch:test:0016", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ac0f4e5388.\\nA means purity >= 70.\\nD means contamination >= 75.\\nC means supervision >= 29.\\nB means traceability >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 97, \\\"supervision\\\": 54, \\\"contamination\\\": 77}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0016:uncertain", "group_id": "wf12:laboratory_batch:test:0016", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ac0f4e5388.\\nA means purity >= 70.\\nD means contamination >= 75.\\nC means supervision >= 29.\\nB means traceability >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 97, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 77}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0017:choice", "group_id": "wf12:laboratory_batch:test:0017", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b2dfda13e2.\\nB means traceability >= 22.\\nA means purity >= 84.\\nC means supervision >= 17.\\nD means contamination >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 18, \\\"supervision\\\": 14, \\\"contamination\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0df24b\": \"Outcome PROCEED.\", \"candidate_d48cc5\": \"Outcome REVIEW.\", \"candidate_e49125\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0df24b\": 0.0, \"candidate_d48cc5\": 0.0, \"candidate_e49125\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0017:score", "group_id": "wf12:laboratory_batch:test:0017", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b2dfda13e2.\\nB means traceability >= 22.\\nA means purity >= 84.\\nC means supervision >= 17.\\nD means contamination >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 18, \\\"supervision\\\": 14, \\\"contamination\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0017:noul", "group_id": "wf12:laboratory_batch:test:0017", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b2dfda13e2.\\nB means traceability >= 22.\\nA means purity >= 84.\\nC means supervision >= 17.\\nD means contamination >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 18, \\\"supervision\\\": 14, \\\"contamination\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0017:uncertain", "group_id": "wf12:laboratory_batch:test:0017", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b2dfda13e2.\\nB means traceability >= 22.\\nA means purity >= 84.\\nC means supervision >= 17.\\nD means contamination >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 61}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.9, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.95, \"true\": 0.050000000000000044}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0018:choice", "group_id": "wf12:laboratory_batch:test:0018", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 582e896060.\\nD means contamination >= 70.\\nC means supervision >= 90.\\nB means traceability >= 71.\\nA means purity >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 38, \\\"supervision\\\": 100, \\\"contamination\\\": 66}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6a0ee2\": \"Outcome PROCEED.\", \"candidate_58d0bb\": \"Outcome REVIEW.\", \"candidate_78abc7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6a0ee2\": 1.0, \"candidate_58d0bb\": 0.0, \"candidate_78abc7\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0018:score", "group_id": "wf12:laboratory_batch:test:0018", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 582e896060.\\nD means contamination >= 70.\\nC means supervision >= 90.\\nB means traceability >= 71.\\nA means purity >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 38, \\\"supervision\\\": 100, \\\"contamination\\\": 66}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0018:noul", "group_id": "wf12:laboratory_batch:test:0018", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 582e896060.\\nD means contamination >= 70.\\nC means supervision >= 90.\\nB means traceability >= 71.\\nA means purity >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 38, \\\"supervision\\\": 100, \\\"contamination\\\": 66}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0018:uncertain", "group_id": "wf12:laboratory_batch:test:0018", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 582e896060.\\nD means contamination >= 70.\\nC means supervision >= 90.\\nB means traceability >= 71.\\nA means purity >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 100, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6a0ee2\": \"Outcome PROCEED.\", \"candidate_58d0bb\": \"Outcome REVIEW.\", \"candidate_78abc7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6a0ee2\": 0.35, \"candidate_58d0bb\": 0.65, \"candidate_78abc7\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0019:choice", "group_id": "wf12:laboratory_batch:test:0019", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0c2657f17a.\\nB means traceability >= 23.\\nA means purity >= 41.\\nD means contamination >= 49.\\nC means supervision >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 41, \\\"traceability\\\": 81, \\\"supervision\\\": 83, \\\"contamination\\\": 83}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8373ba\": \"Outcome PROCEED.\", \"candidate_3b5354\": \"Outcome REVIEW.\", \"candidate_e784be\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8373ba\": 0.0, \"candidate_3b5354\": 1.0, \"candidate_e784be\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0019:score", "group_id": "wf12:laboratory_batch:test:0019", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0c2657f17a.\\nB means traceability >= 23.\\nA means purity >= 41.\\nD means contamination >= 49.\\nC means supervision >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 41, \\\"traceability\\\": 81, \\\"supervision\\\": 83, \\\"contamination\\\": 83}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0019:noul", "group_id": "wf12:laboratory_batch:test:0019", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0c2657f17a.\\nB means traceability >= 23.\\nA means purity >= 41.\\nD means contamination >= 49.\\nC means supervision >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 41, \\\"traceability\\\": 81, \\\"supervision\\\": 83, \\\"contamination\\\": 83}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0019:uncertain", "group_id": "wf12:laboratory_batch:test:0019", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0c2657f17a.\\nB means traceability >= 23.\\nA means purity >= 41.\\nD means contamination >= 49.\\nC means supervision >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 41, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 83}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8373ba\": \"Outcome PROCEED.\", \"candidate_3b5354\": \"Outcome REVIEW.\", \"candidate_e784be\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8373ba\": 0.0, \"candidate_3b5354\": 0.35, \"candidate_e784be\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0020:choice", "group_id": "wf12:laboratory_batch:test:0020", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6ee6771a60.\\nD means contamination >= 40.\\nA means purity >= 46.\\nC means supervision >= 27.\\nB means traceability >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 79, \\\"supervision\\\": 7, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_250f9c\": \"Outcome PROCEED.\", \"candidate_58e0e0\": \"Outcome REVIEW.\", \"candidate_7bc678\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_250f9c\": 0.0, \"candidate_58e0e0\": 0.0, \"candidate_7bc678\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0020:score", "group_id": "wf12:laboratory_batch:test:0020", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6ee6771a60.\\nD means contamination >= 40.\\nA means purity >= 46.\\nC means supervision >= 27.\\nB means traceability >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 79, \\\"supervision\\\": 7, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0020:noul", "group_id": "wf12:laboratory_batch:test:0020", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6ee6771a60.\\nD means contamination >= 40.\\nA means purity >= 46.\\nC means supervision >= 27.\\nB means traceability >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 79, \\\"supervision\\\": 7, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0020:uncertain", "group_id": "wf12:laboratory_batch:test:0020", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6ee6771a60.\\nD means contamination >= 40.\\nA means purity >= 46.\\nC means supervision >= 27.\\nB means traceability >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 79, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.1, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_250f9c\": \"Outcome PROCEED.\", \"candidate_58e0e0\": \"Outcome REVIEW.\", \"candidate_7bc678\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_250f9c\": 0.95, \"candidate_58e0e0\": 0.005000000000000001, \"candidate_7bc678\": 0.045000000000000005}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0021:choice", "group_id": "wf12:laboratory_batch:test:0021", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 358eab5389.\\nA means purity >= 15.\\nC means supervision >= 55.\\nB means traceability >= 35.\\nD means contamination >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 14, \\\"traceability\\\": 21, \\\"supervision\\\": 59, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ab002a\": \"Outcome PROCEED.\", \"candidate_4b4d64\": \"Outcome REVIEW.\", \"candidate_1a379b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ab002a\": 1.0, \"candidate_4b4d64\": 0.0, \"candidate_1a379b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0021:score", "group_id": "wf12:laboratory_batch:test:0021", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 358eab5389.\\nA means purity >= 15.\\nC means supervision >= 55.\\nB means traceability >= 35.\\nD means contamination >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 14, \\\"traceability\\\": 21, \\\"supervision\\\": 59, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0021:noul", "group_id": "wf12:laboratory_batch:test:0021", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 358eab5389.\\nA means purity >= 15.\\nC means supervision >= 55.\\nB means traceability >= 35.\\nD means contamination >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 14, \\\"traceability\\\": 21, \\\"supervision\\\": 59, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0021:uncertain", "group_id": "wf12:laboratory_batch:test:0021", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 358eab5389.\\nA means purity >= 15.\\nC means supervision >= 55.\\nB means traceability >= 35.\\nD means contamination >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 14, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 59, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.65, \"1\": 0.35, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0022:choice", "group_id": "wf12:laboratory_batch:test:0022", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a681b130b8.\\nB means traceability >= 24.\\nA means purity >= 19.\\nD means contamination >= 59.\\nC means supervision >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 68, \\\"traceability\\\": 88, \\\"supervision\\\": 58, \\\"contamination\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fb6702\": \"Outcome PROCEED.\", \"candidate_9746a9\": \"Outcome REVIEW.\", \"candidate_410736\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fb6702\": 0.0, \"candidate_9746a9\": 1.0, \"candidate_410736\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0022:score", "group_id": "wf12:laboratory_batch:test:0022", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a681b130b8.\\nB means traceability >= 24.\\nA means purity >= 19.\\nD means contamination >= 59.\\nC means supervision >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 68, \\\"traceability\\\": 88, \\\"supervision\\\": 58, \\\"contamination\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0022:noul", "group_id": "wf12:laboratory_batch:test:0022", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a681b130b8.\\nB means traceability >= 24.\\nA means purity >= 19.\\nD means contamination >= 59.\\nC means supervision >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 68, \\\"traceability\\\": 88, \\\"supervision\\\": 58, \\\"contamination\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0022:uncertain", "group_id": "wf12:laboratory_batch:test:0022", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a681b130b8.\\nB means traceability >= 24.\\nA means purity >= 19.\\nD means contamination >= 59.\\nC means supervision >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 68, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 84}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.2, \"2\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0023:choice", "group_id": "wf12:laboratory_batch:test:0023", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8b6de1132c.\\nB means traceability >= 88.\\nA means purity >= 79.\\nC means supervision >= 22.\\nD means contamination >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 90, \\\"supervision\\\": 13, \\\"contamination\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_698926\": \"Outcome PROCEED.\", \"candidate_b03396\": \"Outcome REVIEW.\", \"candidate_bef266\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_698926\": 0.0, \"candidate_b03396\": 0.0, \"candidate_bef266\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0023:score", "group_id": "wf12:laboratory_batch:test:0023", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8b6de1132c.\\nB means traceability >= 88.\\nA means purity >= 79.\\nC means supervision >= 22.\\nD means contamination >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 90, \\\"supervision\\\": 13, \\\"contamination\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0023:noul", "group_id": "wf12:laboratory_batch:test:0023", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8b6de1132c.\\nB means traceability >= 88.\\nA means purity >= 79.\\nC means supervision >= 22.\\nD means contamination >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 90, \\\"supervision\\\": 13, \\\"contamination\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0023:uncertain", "group_id": "wf12:laboratory_batch:test:0023", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8b6de1132c.\\nB means traceability >= 88.\\nA means purity >= 79.\\nC means supervision >= 22.\\nD means contamination >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 13, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.095, \"1\": 0.855, \"2\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0024:choice", "group_id": "wf12:laboratory_batch:test:0024", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a5bd9597db.\\nA means purity >= 85.\\nB means traceability >= 22.\\nD means contamination >= 14.\\nC means supervision >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 21, \\\"supervision\\\": 69, \\\"contamination\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_28e22f\": \"Outcome PROCEED.\", \"candidate_9efc96\": \"Outcome REVIEW.\", \"candidate_4f4b6a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_28e22f\": 1.0, \"candidate_9efc96\": 0.0, \"candidate_4f4b6a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0024:score", "group_id": "wf12:laboratory_batch:test:0024", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a5bd9597db.\\nA means purity >= 85.\\nB means traceability >= 22.\\nD means contamination >= 14.\\nC means supervision >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 21, \\\"supervision\\\": 69, \\\"contamination\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0024:noul", "group_id": "wf12:laboratory_batch:test:0024", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a5bd9597db.\\nA means purity >= 85.\\nB means traceability >= 22.\\nD means contamination >= 14.\\nC means supervision >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 21, \\\"supervision\\\": 69, \\\"contamination\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0024:uncertain", "group_id": "wf12:laboratory_batch:test:0024", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a5bd9597db.\\nA means purity >= 85.\\nB means traceability >= 22.\\nD means contamination >= 14.\\nC means supervision >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.65, \\\"C\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.96, \"true\": 0.039999999999999994}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0025:choice", "group_id": "wf12:laboratory_batch:test:0025", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference daaa35f907.\\nC means supervision >= 75.\\nA means purity >= 51.\\nB means traceability >= 49.\\nD means contamination >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 58, \\\"supervision\\\": 31, \\\"contamination\\\": 43}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8baeb9\": \"Outcome PROCEED.\", \"candidate_2f12a1\": \"Outcome REVIEW.\", \"candidate_23f57d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8baeb9\": 0.0, \"candidate_2f12a1\": 1.0, \"candidate_23f57d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0025:score", "group_id": "wf12:laboratory_batch:test:0025", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference daaa35f907.\\nC means supervision >= 75.\\nA means purity >= 51.\\nB means traceability >= 49.\\nD means contamination >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 58, \\\"supervision\\\": 31, \\\"contamination\\\": 43}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0025:noul", "group_id": "wf12:laboratory_batch:test:0025", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference daaa35f907.\\nC means supervision >= 75.\\nA means purity >= 51.\\nB means traceability >= 49.\\nD means contamination >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 58, \\\"supervision\\\": 31, \\\"contamination\\\": 43}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0025:uncertain", "group_id": "wf12:laboratory_batch:test:0025", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference daaa35f907.\\nC means supervision >= 75.\\nA means purity >= 51.\\nB means traceability >= 49.\\nD means contamination >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 31, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.65, \"true\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0026:choice", "group_id": "wf12:laboratory_batch:test:0026", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 39b1b160c2.\\nA means purity >= 70.\\nC means supervision >= 20.\\nD means contamination >= 40.\\nB means traceability >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 0, \\\"traceability\\\": 18, \\\"supervision\\\": 0, \\\"contamination\\\": 53}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_62b9b5\": \"Outcome PROCEED.\", \"candidate_b8d58c\": \"Outcome REVIEW.\", \"candidate_c8b9f3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_62b9b5\": 0.0, \"candidate_b8d58c\": 0.0, \"candidate_c8b9f3\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0026:score", "group_id": "wf12:laboratory_batch:test:0026", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 39b1b160c2.\\nA means purity >= 70.\\nC means supervision >= 20.\\nD means contamination >= 40.\\nB means traceability >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 0, \\\"traceability\\\": 18, \\\"supervision\\\": 0, \\\"contamination\\\": 53}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0026:noul", "group_id": "wf12:laboratory_batch:test:0026", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 39b1b160c2.\\nA means purity >= 70.\\nC means supervision >= 20.\\nD means contamination >= 40.\\nB means traceability >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 0, \\\"traceability\\\": 18, \\\"supervision\\\": 0, \\\"contamination\\\": 53}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0026:uncertain", "group_id": "wf12:laboratory_batch:test:0026", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 39b1b160c2.\\nA means purity >= 70.\\nC means supervision >= 20.\\nD means contamination >= 40.\\nB means traceability >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 18, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 53}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"C\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9, \"true\": 0.09999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0027:choice", "group_id": "wf12:laboratory_batch:test:0027", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6071412dfe.\\nA means purity >= 31.\\nD means contamination >= 77.\\nC means supervision >= 32.\\nB means traceability >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 12, \\\"supervision\\\": 95, \\\"contamination\\\": 62}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_da7bfc\": \"Outcome PROCEED.\", \"candidate_60fbe9\": \"Outcome REVIEW.\", \"candidate_555936\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_da7bfc\": 1.0, \"candidate_60fbe9\": 0.0, \"candidate_555936\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0027:score", "group_id": "wf12:laboratory_batch:test:0027", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6071412dfe.\\nA means purity >= 31.\\nD means contamination >= 77.\\nC means supervision >= 32.\\nB means traceability >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 12, \\\"supervision\\\": 95, \\\"contamination\\\": 62}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0027:noul", "group_id": "wf12:laboratory_batch:test:0027", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6071412dfe.\\nA means purity >= 31.\\nD means contamination >= 77.\\nC means supervision >= 32.\\nB means traceability >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 12, \\\"supervision\\\": 95, \\\"contamination\\\": 62}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0027:uncertain", "group_id": "wf12:laboratory_batch:test:0027", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6071412dfe.\\nA means purity >= 31.\\nD means contamination >= 77.\\nC means supervision >= 32.\\nB means traceability >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 12, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_da7bfc\": \"Outcome PROCEED.\", \"candidate_60fbe9\": \"Outcome REVIEW.\", \"candidate_555936\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_da7bfc\": 0.6400000000000001, \"candidate_60fbe9\": 0.32, \"candidate_555936\": 0.039999999999999994}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0028:choice", "group_id": "wf12:laboratory_batch:test:0028", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c5389166f4.\\nA means purity >= 77.\\nC means supervision >= 90.\\nD means contamination >= 73.\\nB means traceability >= 20.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 4, \\\"supervision\\\": 97, \\\"contamination\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_321c12\": \"Outcome PROCEED.\", \"candidate_6f5f17\": \"Outcome REVIEW.\", \"candidate_239d56\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_321c12\": 0.0, \"candidate_6f5f17\": 1.0, \"candidate_239d56\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0028:score", "group_id": "wf12:laboratory_batch:test:0028", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c5389166f4.\\nA means purity >= 77.\\nC means supervision >= 90.\\nD means contamination >= 73.\\nB means traceability >= 20.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 4, \\\"supervision\\\": 97, \\\"contamination\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0028:noul", "group_id": "wf12:laboratory_batch:test:0028", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c5389166f4.\\nA means purity >= 77.\\nC means supervision >= 90.\\nD means contamination >= 73.\\nB means traceability >= 20.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 4, \\\"supervision\\\": 97, \\\"contamination\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0028:uncertain", "group_id": "wf12:laboratory_batch:test:0028", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c5389166f4.\\nA means purity >= 77.\\nC means supervision >= 90.\\nD means contamination >= 73.\\nB means traceability >= 20.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 76}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_321c12\": \"Outcome PROCEED.\", \"candidate_6f5f17\": \"Outcome REVIEW.\", \"candidate_239d56\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_321c12\": 0.0, \"candidate_6f5f17\": 0.2, \"candidate_239d56\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0029:choice", "group_id": "wf12:laboratory_batch:test:0029", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 43a32d0608.\\nB means traceability >= 17.\\nD means contamination >= 39.\\nA means purity >= 14.\\nC means supervision >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": 66, \\\"supervision\\\": 4, \\\"contamination\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f879c2\": \"Outcome PROCEED.\", \"candidate_b527ff\": \"Outcome REVIEW.\", \"candidate_2ee117\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f879c2\": 0.0, \"candidate_b527ff\": 0.0, \"candidate_2ee117\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0029:score", "group_id": "wf12:laboratory_batch:test:0029", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 43a32d0608.\\nB means traceability >= 17.\\nD means contamination >= 39.\\nA means purity >= 14.\\nC means supervision >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": 66, \\\"supervision\\\": 4, \\\"contamination\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0029:noul", "group_id": "wf12:laboratory_batch:test:0029", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 43a32d0608.\\nB means traceability >= 17.\\nD means contamination >= 39.\\nA means purity >= 14.\\nC means supervision >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": 66, \\\"supervision\\\": 4, \\\"contamination\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0029:uncertain", "group_id": "wf12:laboratory_batch:test:0029", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 43a32d0608.\\nB means traceability >= 17.\\nD means contamination >= 39.\\nA means purity >= 14.\\nC means supervision >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 4, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.9, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f879c2\": \"Outcome PROCEED.\", \"candidate_b527ff\": \"Outcome REVIEW.\", \"candidate_2ee117\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f879c2\": 0.0, \"candidate_b527ff\": 0.050000000000000044, \"candidate_2ee117\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0030:choice", "group_id": "wf12:laboratory_batch:test:0030", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0b1dc72a31.\\nA means purity >= 18.\\nC means supervision >= 23.\\nB means traceability >= 12.\\nD means contamination >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 38, \\\"supervision\\\": 65, \\\"contamination\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_62e4aa\": \"Outcome PROCEED.\", \"candidate_8f5e16\": \"Outcome REVIEW.\", \"candidate_59ef2c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_62e4aa\": 1.0, \"candidate_8f5e16\": 0.0, \"candidate_59ef2c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0030:score", "group_id": "wf12:laboratory_batch:test:0030", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0b1dc72a31.\\nA means purity >= 18.\\nC means supervision >= 23.\\nB means traceability >= 12.\\nD means contamination >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 38, \\\"supervision\\\": 65, \\\"contamination\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0030:noul", "group_id": "wf12:laboratory_batch:test:0030", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0b1dc72a31.\\nA means purity >= 18.\\nC means supervision >= 23.\\nB means traceability >= 12.\\nD means contamination >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 38, \\\"supervision\\\": 65, \\\"contamination\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0030:uncertain", "group_id": "wf12:laboratory_batch:test:0030", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0b1dc72a31.\\nA means purity >= 18.\\nC means supervision >= 23.\\nB means traceability >= 12.\\nD means contamination >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.2, \\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.4316000000000001, \"1\": 0.4459, \"2\": 0.12250000000000003}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0031:choice", "group_id": "wf12:laboratory_batch:test:0031", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5f02719a24.\\nA means purity >= 36.\\nB means traceability >= 11.\\nC means supervision >= 65.\\nD means contamination >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": 6, \\\"supervision\\\": 25, \\\"contamination\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0ef0db\": \"Outcome PROCEED.\", \"candidate_78d0e7\": \"Outcome REVIEW.\", \"candidate_b3673b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0ef0db\": 0.0, \"candidate_78d0e7\": 1.0, \"candidate_b3673b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0031:score", "group_id": "wf12:laboratory_batch:test:0031", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5f02719a24.\\nA means purity >= 36.\\nB means traceability >= 11.\\nC means supervision >= 65.\\nD means contamination >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": 6, \\\"supervision\\\": 25, \\\"contamination\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0031:noul", "group_id": "wf12:laboratory_batch:test:0031", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5f02719a24.\\nA means purity >= 36.\\nB means traceability >= 11.\\nC means supervision >= 65.\\nD means contamination >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": 6, \\\"supervision\\\": 25, \\\"contamination\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0031:uncertain", "group_id": "wf12:laboratory_batch:test:0031", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5f02719a24.\\nA means purity >= 36.\\nB means traceability >= 11.\\nC means supervision >= 65.\\nD means contamination >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 41}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9073972602739726, \"1\": 0.0926027397260274, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0032:choice", "group_id": "wf12:laboratory_batch:test:0032", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 15e302bee9.\\nA means purity >= 37.\\nB means traceability >= 57.\\nD means contamination >= 26.\\nC means supervision >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 76, \\\"traceability\\\": 81, \\\"supervision\\\": 5, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_92f951\": \"Outcome PROCEED.\", \"candidate_23cb43\": \"Outcome REVIEW.\", \"candidate_5ed6b2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_92f951\": 0.0, \"candidate_23cb43\": 0.0, \"candidate_5ed6b2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0032:score", "group_id": "wf12:laboratory_batch:test:0032", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 15e302bee9.\\nA means purity >= 37.\\nB means traceability >= 57.\\nD means contamination >= 26.\\nC means supervision >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 76, \\\"traceability\\\": 81, \\\"supervision\\\": 5, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0032:noul", "group_id": "wf12:laboratory_batch:test:0032", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 15e302bee9.\\nA means purity >= 37.\\nB means traceability >= 57.\\nD means contamination >= 26.\\nC means supervision >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 76, \\\"traceability\\\": 81, \\\"supervision\\\": 5, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0032:uncertain", "group_id": "wf12:laboratory_batch:test:0032", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 15e302bee9.\\nA means purity >= 37.\\nB means traceability >= 57.\\nD means contamination >= 26.\\nC means supervision >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 76, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 99}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.1, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0033:choice", "group_id": "wf12:laboratory_batch:test:0033", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 929c1ac227.\\nD means contamination >= 24.\\nC means supervision >= 30.\\nA means purity >= 85.\\nB means traceability >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 21, \\\"supervision\\\": 36, \\\"contamination\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f206e8\": \"Outcome PROCEED.\", \"candidate_60399f\": \"Outcome REVIEW.\", \"candidate_ca36fc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f206e8\": 1.0, \"candidate_60399f\": 0.0, \"candidate_ca36fc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0033:score", "group_id": "wf12:laboratory_batch:test:0033", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 929c1ac227.\\nD means contamination >= 24.\\nC means supervision >= 30.\\nA means purity >= 85.\\nB means traceability >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 21, \\\"supervision\\\": 36, \\\"contamination\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0033:noul", "group_id": "wf12:laboratory_batch:test:0033", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 929c1ac227.\\nD means contamination >= 24.\\nC means supervision >= 30.\\nA means purity >= 85.\\nB means traceability >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 21, \\\"supervision\\\": 36, \\\"contamination\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0033:uncertain", "group_id": "wf12:laboratory_batch:test:0033", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 929c1ac227.\\nD means contamination >= 24.\\nC means supervision >= 30.\\nA means purity >= 85.\\nB means traceability >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.2, \\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.93, \"true\": 0.06999999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0034:choice", "group_id": "wf12:laboratory_batch:test:0034", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a4593343df.\\nC means supervision >= 51.\\nA means purity >= 34.\\nD means contamination >= 16.\\nB means traceability >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 99, \\\"supervision\\\": 77, \\\"contamination\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c9d59c\": \"Outcome PROCEED.\", \"candidate_674cd2\": \"Outcome REVIEW.\", \"candidate_d6bd21\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c9d59c\": 0.0, \"candidate_674cd2\": 1.0, \"candidate_d6bd21\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0034:score", "group_id": "wf12:laboratory_batch:test:0034", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a4593343df.\\nC means supervision >= 51.\\nA means purity >= 34.\\nD means contamination >= 16.\\nB means traceability >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 99, \\\"supervision\\\": 77, \\\"contamination\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0034:noul", "group_id": "wf12:laboratory_batch:test:0034", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a4593343df.\\nC means supervision >= 51.\\nA means purity >= 34.\\nD means contamination >= 16.\\nB means traceability >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 99, \\\"supervision\\\": 77, \\\"contamination\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0034:uncertain", "group_id": "wf12:laboratory_batch:test:0034", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a4593343df.\\nC means supervision >= 51.\\nA means purity >= 34.\\nD means contamination >= 16.\\nB means traceability >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 55}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"C\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.65, \"true\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0035:choice", "group_id": "wf12:laboratory_batch:test:0035", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a68d8cfc17.\\nA means purity >= 12.\\nB means traceability >= 84.\\nD means contamination >= 18.\\nC means supervision >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": 86, \\\"supervision\\\": 17, \\\"contamination\\\": 92}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_01eb95\": \"Outcome PROCEED.\", \"candidate_b3ce9e\": \"Outcome REVIEW.\", \"candidate_f01068\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_01eb95\": 0.0, \"candidate_b3ce9e\": 0.0, \"candidate_f01068\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0035:score", "group_id": "wf12:laboratory_batch:test:0035", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a68d8cfc17.\\nA means purity >= 12.\\nB means traceability >= 84.\\nD means contamination >= 18.\\nC means supervision >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": 86, \\\"supervision\\\": 17, \\\"contamination\\\": 92}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0035:noul", "group_id": "wf12:laboratory_batch:test:0035", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a68d8cfc17.\\nA means purity >= 12.\\nB means traceability >= 84.\\nD means contamination >= 18.\\nC means supervision >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": 86, \\\"supervision\\\": 17, \\\"contamination\\\": 92}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0035:uncertain", "group_id": "wf12:laboratory_batch:test:0035", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a68d8cfc17.\\nA means purity >= 12.\\nB means traceability >= 84.\\nD means contamination >= 18.\\nC means supervision >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 92}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09999999999999998, \"true\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0036:choice", "group_id": "wf12:laboratory_batch:test:0036", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 82ce32ba3b.\\nC means supervision >= 10.\\nA means purity >= 58.\\nD means contamination >= 25.\\nB means traceability >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 63, \\\"traceability\\\": 79, \\\"supervision\\\": 13, \\\"contamination\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b6e013\": \"Outcome PROCEED.\", \"candidate_4ddf39\": \"Outcome REVIEW.\", \"candidate_86a2e7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b6e013\": 1.0, \"candidate_4ddf39\": 0.0, \"candidate_86a2e7\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0036:score", "group_id": "wf12:laboratory_batch:test:0036", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 82ce32ba3b.\\nC means supervision >= 10.\\nA means purity >= 58.\\nD means contamination >= 25.\\nB means traceability >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 63, \\\"traceability\\\": 79, \\\"supervision\\\": 13, \\\"contamination\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0036:noul", "group_id": "wf12:laboratory_batch:test:0036", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 82ce32ba3b.\\nC means supervision >= 10.\\nA means purity >= 58.\\nD means contamination >= 25.\\nB means traceability >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 63, \\\"traceability\\\": 79, \\\"supervision\\\": 13, \\\"contamination\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0036:uncertain", "group_id": "wf12:laboratory_batch:test:0036", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 82ce32ba3b.\\nC means supervision >= 10.\\nA means purity >= 58.\\nD means contamination >= 25.\\nB means traceability >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 63, \\\"traceability\\\": 79, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.2, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b6e013\": \"Outcome PROCEED.\", \"candidate_4ddf39\": \"Outcome REVIEW.\", \"candidate_86a2e7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b6e013\": 0.8000000000000002, \"candidate_4ddf39\": 0.04000000000000001, \"candidate_86a2e7\": 0.16000000000000003}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0037:choice", "group_id": "wf12:laboratory_batch:test:0037", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a80588b1af.\\nD means contamination >= 31.\\nA means purity >= 44.\\nB means traceability >= 39.\\nC means supervision >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 21, \\\"traceability\\\": 28, \\\"supervision\\\": 6, \\\"contamination\\\": 24}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_15cee8\": \"Outcome PROCEED.\", \"candidate_8f0df7\": \"Outcome REVIEW.\", \"candidate_0f74ca\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_15cee8\": 0.0, \"candidate_8f0df7\": 1.0, \"candidate_0f74ca\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0037:score", "group_id": "wf12:laboratory_batch:test:0037", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a80588b1af.\\nD means contamination >= 31.\\nA means purity >= 44.\\nB means traceability >= 39.\\nC means supervision >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 21, \\\"traceability\\\": 28, \\\"supervision\\\": 6, \\\"contamination\\\": 24}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0037:noul", "group_id": "wf12:laboratory_batch:test:0037", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a80588b1af.\\nD means contamination >= 31.\\nA means purity >= 44.\\nB means traceability >= 39.\\nC means supervision >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 21, \\\"traceability\\\": 28, \\\"supervision\\\": 6, \\\"contamination\\\": 24}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0037:uncertain", "group_id": "wf12:laboratory_batch:test:0037", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a80588b1af.\\nD means contamination >= 31.\\nA means purity >= 44.\\nB means traceability >= 39.\\nC means supervision >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 28, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 24}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_15cee8\": \"Outcome PROCEED.\", \"candidate_8f0df7\": \"Outcome REVIEW.\", \"candidate_0f74ca\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_15cee8\": 0.8, \"candidate_8f0df7\": 0.19999999999999996, \"candidate_0f74ca\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0038:choice", "group_id": "wf12:laboratory_batch:test:0038", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 11944dddda.\\nD means contamination >= 31.\\nA means purity >= 21.\\nB means traceability >= 28.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 59, \\\"traceability\\\": 85, \\\"supervision\\\": 34, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8a3d80\": \"Outcome PROCEED.\", \"candidate_a2fc0e\": \"Outcome REVIEW.\", \"candidate_1fb3c7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8a3d80\": 0.0, \"candidate_a2fc0e\": 0.0, \"candidate_1fb3c7\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0038:score", "group_id": "wf12:laboratory_batch:test:0038", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 11944dddda.\\nD means contamination >= 31.\\nA means purity >= 21.\\nB means traceability >= 28.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 59, \\\"traceability\\\": 85, \\\"supervision\\\": 34, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0038:noul", "group_id": "wf12:laboratory_batch:test:0038", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 11944dddda.\\nD means contamination >= 31.\\nA means purity >= 21.\\nB means traceability >= 28.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 59, \\\"traceability\\\": 85, \\\"supervision\\\": 34, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0038:uncertain", "group_id": "wf12:laboratory_batch:test:0038", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 11944dddda.\\nD means contamination >= 31.\\nA means purity >= 21.\\nB means traceability >= 28.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 85, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 98}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8a3d80\": \"Outcome PROCEED.\", \"candidate_a2fc0e\": \"Outcome REVIEW.\", \"candidate_1fb3c7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8a3d80\": 0.0, \"candidate_a2fc0e\": 0.05, \"candidate_1fb3c7\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0039:choice", "group_id": "wf12:laboratory_batch:test:0039", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f4290f1e24.\\nB means traceability >= 69.\\nD means contamination >= 60.\\nA means purity >= 65.\\nC means supervision >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 24, \\\"traceability\\\": 21, \\\"supervision\\\": 29, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d3b04b\": \"Outcome PROCEED.\", \"candidate_a5d07e\": \"Outcome REVIEW.\", \"candidate_be61a2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d3b04b\": 1.0, \"candidate_a5d07e\": 0.0, \"candidate_be61a2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0039:score", "group_id": "wf12:laboratory_batch:test:0039", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f4290f1e24.\\nB means traceability >= 69.\\nD means contamination >= 60.\\nA means purity >= 65.\\nC means supervision >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 24, \\\"traceability\\\": 21, \\\"supervision\\\": 29, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0039:noul", "group_id": "wf12:laboratory_batch:test:0039", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f4290f1e24.\\nB means traceability >= 69.\\nD means contamination >= 60.\\nA means purity >= 65.\\nC means supervision >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 24, \\\"traceability\\\": 21, \\\"supervision\\\": 29, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0039:uncertain", "group_id": "wf12:laboratory_batch:test:0039", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f4290f1e24.\\nB means traceability >= 69.\\nD means contamination >= 60.\\nA means purity >= 65.\\nC means supervision >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 24, \\\"traceability\\\": 21, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.22749999999999998, \"1\": 0.545, \"2\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0040:choice", "group_id": "wf12:laboratory_batch:test:0040", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 92a59be8e5.\\nC means supervision >= 71.\\nB means traceability >= 69.\\nD means contamination >= 76.\\nA means purity >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 79, \\\"traceability\\\": 26, \\\"supervision\\\": 95, \\\"contamination\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a6b9d\": \"Outcome PROCEED.\", \"candidate_cb401f\": \"Outcome REVIEW.\", \"candidate_4d5acc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a6b9d\": 0.0, \"candidate_cb401f\": 1.0, \"candidate_4d5acc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0040:score", "group_id": "wf12:laboratory_batch:test:0040", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 92a59be8e5.\\nC means supervision >= 71.\\nB means traceability >= 69.\\nD means contamination >= 76.\\nA means purity >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 79, \\\"traceability\\\": 26, \\\"supervision\\\": 95, \\\"contamination\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0040:noul", "group_id": "wf12:laboratory_batch:test:0040", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 92a59be8e5.\\nC means supervision >= 71.\\nB means traceability >= 69.\\nD means contamination >= 76.\\nA means purity >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 79, \\\"traceability\\\": 26, \\\"supervision\\\": 95, \\\"contamination\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0040:uncertain", "group_id": "wf12:laboratory_batch:test:0040", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 92a59be8e5.\\nC means supervision >= 71.\\nB means traceability >= 69.\\nD means contamination >= 76.\\nA means purity >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 26, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 79}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.8, \"2\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0041:choice", "group_id": "wf12:laboratory_batch:test:0041", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 43c9190901.\\nA means purity >= 56.\\nB means traceability >= 13.\\nC means supervision >= 38.\\nD means contamination >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 22, \\\"traceability\\\": 68, \\\"supervision\\\": 30, \\\"contamination\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6a9f1c\": \"Outcome PROCEED.\", \"candidate_e03bd5\": \"Outcome REVIEW.\", \"candidate_7f17f4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6a9f1c\": 0.0, \"candidate_e03bd5\": 0.0, \"candidate_7f17f4\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0041:score", "group_id": "wf12:laboratory_batch:test:0041", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 43c9190901.\\nA means purity >= 56.\\nB means traceability >= 13.\\nC means supervision >= 38.\\nD means contamination >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 22, \\\"traceability\\\": 68, \\\"supervision\\\": 30, \\\"contamination\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0041:noul", "group_id": "wf12:laboratory_batch:test:0041", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 43c9190901.\\nA means purity >= 56.\\nB means traceability >= 13.\\nC means supervision >= 38.\\nD means contamination >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 22, \\\"traceability\\\": 68, \\\"supervision\\\": 30, \\\"contamination\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0041:uncertain", "group_id": "wf12:laboratory_batch:test:0041", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 43c9190901.\\nA means purity >= 56.\\nB means traceability >= 13.\\nC means supervision >= 38.\\nD means contamination >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.95, \\\"C\\\": 0.05, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.014024999999999998, \"1\": 0.13097499999999998, \"2\": 0.855}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0042:choice", "group_id": "wf12:laboratory_batch:test:0042", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8195bda56a.\\nD means contamination >= 11.\\nC means supervision >= 50.\\nB means traceability >= 56.\\nA means purity >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 30, \\\"supervision\\\": 76, \\\"contamination\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dd0649\": \"Outcome PROCEED.\", \"candidate_7645cb\": \"Outcome REVIEW.\", \"candidate_bdbcd3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dd0649\": 1.0, \"candidate_7645cb\": 0.0, \"candidate_bdbcd3\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0042:score", "group_id": "wf12:laboratory_batch:test:0042", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8195bda56a.\\nD means contamination >= 11.\\nC means supervision >= 50.\\nB means traceability >= 56.\\nA means purity >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 30, \\\"supervision\\\": 76, \\\"contamination\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0042:noul", "group_id": "wf12:laboratory_batch:test:0042", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8195bda56a.\\nD means contamination >= 11.\\nC means supervision >= 50.\\nB means traceability >= 56.\\nA means purity >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 30, \\\"supervision\\\": 76, \\\"contamination\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0042:uncertain", "group_id": "wf12:laboratory_batch:test:0042", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8195bda56a.\\nD means contamination >= 11.\\nC means supervision >= 50.\\nB means traceability >= 56.\\nA means purity >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.8, \\\"C\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.72, \"true\": 0.28}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0043:choice", "group_id": "wf12:laboratory_batch:test:0043", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e68c415ed.\\nB means traceability >= 59.\\nA means purity >= 89.\\nC means supervision >= 62.\\nD means contamination >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 21, \\\"supervision\\\": 3, \\\"contamination\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b6af38\": \"Outcome PROCEED.\", \"candidate_bc1b93\": \"Outcome REVIEW.\", \"candidate_ca460b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b6af38\": 0.0, \"candidate_bc1b93\": 1.0, \"candidate_ca460b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0043:score", "group_id": "wf12:laboratory_batch:test:0043", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e68c415ed.\\nB means traceability >= 59.\\nA means purity >= 89.\\nC means supervision >= 62.\\nD means contamination >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 21, \\\"supervision\\\": 3, \\\"contamination\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0043:noul", "group_id": "wf12:laboratory_batch:test:0043", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e68c415ed.\\nB means traceability >= 59.\\nA means purity >= 89.\\nC means supervision >= 62.\\nD means contamination >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 21, \\\"supervision\\\": 3, \\\"contamination\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0043:uncertain", "group_id": "wf12:laboratory_batch:test:0043", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e68c415ed.\\nB means traceability >= 59.\\nA means purity >= 89.\\nC means supervision >= 62.\\nD means contamination >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.2, \\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.7725, \"true\": 0.22750000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0044:choice", "group_id": "wf12:laboratory_batch:test:0044", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7aa0b96ac7.\\nA means purity >= 67.\\nD means contamination >= 61.\\nB means traceability >= 55.\\nC means supervision >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 59, \\\"traceability\\\": 61, \\\"supervision\\\": 25, \\\"contamination\\\": 87}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a98adf\": \"Outcome PROCEED.\", \"candidate_90f3e1\": \"Outcome REVIEW.\", \"candidate_d24501\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a98adf\": 0.0, \"candidate_90f3e1\": 0.0, \"candidate_d24501\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0044:score", "group_id": "wf12:laboratory_batch:test:0044", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7aa0b96ac7.\\nA means purity >= 67.\\nD means contamination >= 61.\\nB means traceability >= 55.\\nC means supervision >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 59, \\\"traceability\\\": 61, \\\"supervision\\\": 25, \\\"contamination\\\": 87}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0044:noul", "group_id": "wf12:laboratory_batch:test:0044", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7aa0b96ac7.\\nA means purity >= 67.\\nD means contamination >= 61.\\nB means traceability >= 55.\\nC means supervision >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 59, \\\"traceability\\\": 61, \\\"supervision\\\": 25, \\\"contamination\\\": 87}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0044:uncertain", "group_id": "wf12:laboratory_batch:test:0044", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7aa0b96ac7.\\nA means purity >= 67.\\nD means contamination >= 61.\\nB means traceability >= 55.\\nC means supervision >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 61, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 87}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09999999999999998, \"true\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0045:choice", "group_id": "wf12:laboratory_batch:test:0045", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0b80ca215a.\\nC means supervision >= 72.\\nA means purity >= 30.\\nD means contamination >= 31.\\nB means traceability >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 79, \\\"traceability\\\": 76, \\\"supervision\\\": 87, \\\"contamination\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_35f770\": \"Outcome PROCEED.\", \"candidate_98db02\": \"Outcome REVIEW.\", \"candidate_fc4916\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_35f770\": 1.0, \"candidate_98db02\": 0.0, \"candidate_fc4916\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0045:score", "group_id": "wf12:laboratory_batch:test:0045", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0b80ca215a.\\nC means supervision >= 72.\\nA means purity >= 30.\\nD means contamination >= 31.\\nB means traceability >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 79, \\\"traceability\\\": 76, \\\"supervision\\\": 87, \\\"contamination\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0045:noul", "group_id": "wf12:laboratory_batch:test:0045", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0b80ca215a.\\nC means supervision >= 72.\\nA means purity >= 30.\\nD means contamination >= 31.\\nB means traceability >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 79, \\\"traceability\\\": 76, \\\"supervision\\\": 87, \\\"contamination\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0045:uncertain", "group_id": "wf12:laboratory_batch:test:0045", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0b80ca215a.\\nC means supervision >= 72.\\nA means purity >= 30.\\nD means contamination >= 31.\\nB means traceability >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 79, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 87, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_35f770\": \"Outcome PROCEED.\", \"candidate_98db02\": \"Outcome REVIEW.\", \"candidate_fc4916\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_35f770\": 0.65, \"candidate_98db02\": 0.35, \"candidate_fc4916\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0046:choice", "group_id": "wf12:laboratory_batch:test:0046", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d7bc70883c.\\nA means purity >= 15.\\nB means traceability >= 14.\\nC means supervision >= 42.\\nD means contamination >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 13, \\\"traceability\\\": 62, \\\"supervision\\\": 1, \\\"contamination\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b87a59\": \"Outcome PROCEED.\", \"candidate_f046e1\": \"Outcome REVIEW.\", \"candidate_6d9746\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b87a59\": 0.0, \"candidate_f046e1\": 1.0, \"candidate_6d9746\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0046:score", "group_id": "wf12:laboratory_batch:test:0046", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d7bc70883c.\\nA means purity >= 15.\\nB means traceability >= 14.\\nC means supervision >= 42.\\nD means contamination >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 13, \\\"traceability\\\": 62, \\\"supervision\\\": 1, \\\"contamination\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0046:noul", "group_id": "wf12:laboratory_batch:test:0046", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d7bc70883c.\\nA means purity >= 15.\\nB means traceability >= 14.\\nC means supervision >= 42.\\nD means contamination >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 13, \\\"traceability\\\": 62, \\\"supervision\\\": 1, \\\"contamination\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0046:uncertain", "group_id": "wf12:laboratory_batch:test:0046", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d7bc70883c.\\nA means purity >= 15.\\nB means traceability >= 14.\\nC means supervision >= 42.\\nD means contamination >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 13, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 1, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b87a59\": \"Outcome PROCEED.\", \"candidate_f046e1\": \"Outcome REVIEW.\", \"candidate_6d9746\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b87a59\": 0.0, \"candidate_f046e1\": 0.8, \"candidate_6d9746\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0047:choice", "group_id": "wf12:laboratory_batch:test:0047", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fcf4cebfc6.\\nB means traceability >= 17.\\nD means contamination >= 55.\\nA means purity >= 79.\\nC means supervision >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 98, \\\"traceability\\\": 2, \\\"supervision\\\": 20, \\\"contamination\\\": 73}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5a8a6e\": \"Outcome PROCEED.\", \"candidate_0b697a\": \"Outcome REVIEW.\", \"candidate_a91df8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5a8a6e\": 0.0, \"candidate_0b697a\": 0.0, \"candidate_a91df8\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0047:score", "group_id": "wf12:laboratory_batch:test:0047", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fcf4cebfc6.\\nB means traceability >= 17.\\nD means contamination >= 55.\\nA means purity >= 79.\\nC means supervision >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 98, \\\"traceability\\\": 2, \\\"supervision\\\": 20, \\\"contamination\\\": 73}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0047:noul", "group_id": "wf12:laboratory_batch:test:0047", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fcf4cebfc6.\\nB means traceability >= 17.\\nD means contamination >= 55.\\nA means purity >= 79.\\nC means supervision >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 98, \\\"traceability\\\": 2, \\\"supervision\\\": 20, \\\"contamination\\\": 73}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0047:uncertain", "group_id": "wf12:laboratory_batch:test:0047", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fcf4cebfc6.\\nB means traceability >= 17.\\nD means contamination >= 55.\\nA means purity >= 79.\\nC means supervision >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"B\\\": 0.9, \\\"C\\\": 0.95, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5a8a6e\": \"Outcome PROCEED.\", \"candidate_0b697a\": \"Outcome REVIEW.\", \"candidate_a91df8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5a8a6e\": 0.04952500000000004, \"candidate_0b697a\": 0.9029749999999999, \"candidate_a91df8\": 0.047500000000000035}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0048:choice", "group_id": "wf12:laboratory_batch:test:0048", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 332696ac90.\\nD means contamination >= 53.\\nC means supervision >= 74.\\nA means purity >= 75.\\nB means traceability >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 63, \\\"supervision\\\": 100, \\\"contamination\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_983620\": \"Outcome PROCEED.\", \"candidate_0eccdf\": \"Outcome REVIEW.\", \"candidate_7cf36d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_983620\": 1.0, \"candidate_0eccdf\": 0.0, \"candidate_7cf36d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0048:score", "group_id": "wf12:laboratory_batch:test:0048", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 332696ac90.\\nD means contamination >= 53.\\nC means supervision >= 74.\\nA means purity >= 75.\\nB means traceability >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 63, \\\"supervision\\\": 100, \\\"contamination\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0048:noul", "group_id": "wf12:laboratory_batch:test:0048", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 332696ac90.\\nD means contamination >= 53.\\nC means supervision >= 74.\\nA means purity >= 75.\\nB means traceability >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 63, \\\"supervision\\\": 100, \\\"contamination\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0048:uncertain", "group_id": "wf12:laboratory_batch:test:0048", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 332696ac90.\\nD means contamination >= 53.\\nC means supervision >= 74.\\nA means purity >= 75.\\nB means traceability >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 100, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.19999999999999996, \"1\": 0.8, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0049:choice", "group_id": "wf12:laboratory_batch:test:0049", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b722be829c.\\nC means supervision >= 69.\\nA means purity >= 59.\\nD means contamination >= 43.\\nB means traceability >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 49, \\\"traceability\\\": 14, \\\"supervision\\\": 43, \\\"contamination\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5198c9\": \"Outcome PROCEED.\", \"candidate_78793e\": \"Outcome REVIEW.\", \"candidate_618d9b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5198c9\": 0.0, \"candidate_78793e\": 1.0, \"candidate_618d9b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0049:score", "group_id": "wf12:laboratory_batch:test:0049", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b722be829c.\\nC means supervision >= 69.\\nA means purity >= 59.\\nD means contamination >= 43.\\nB means traceability >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 49, \\\"traceability\\\": 14, \\\"supervision\\\": 43, \\\"contamination\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0049:noul", "group_id": "wf12:laboratory_batch:test:0049", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b722be829c.\\nC means supervision >= 69.\\nA means purity >= 59.\\nD means contamination >= 43.\\nB means traceability >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 49, \\\"traceability\\\": 14, \\\"supervision\\\": 43, \\\"contamination\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0049:uncertain", "group_id": "wf12:laboratory_batch:test:0049", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b722be829c.\\nC means supervision >= 69.\\nA means purity >= 59.\\nD means contamination >= 43.\\nB means traceability >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 14, \\\"supervision\\\": 43, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.8, \"2\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0050:choice", "group_id": "wf12:laboratory_batch:test:0050", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0a4a5287c9.\\nB means traceability >= 72.\\nD means contamination >= 28.\\nA means purity >= 19.\\nC means supervision >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 62, \\\"traceability\\\": 77, \\\"supervision\\\": 78, \\\"contamination\\\": 45}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_293c95\": \"Outcome PROCEED.\", \"candidate_ec5233\": \"Outcome REVIEW.\", \"candidate_00ee40\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_293c95\": 0.0, \"candidate_ec5233\": 0.0, \"candidate_00ee40\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0050:score", "group_id": "wf12:laboratory_batch:test:0050", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0a4a5287c9.\\nB means traceability >= 72.\\nD means contamination >= 28.\\nA means purity >= 19.\\nC means supervision >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 62, \\\"traceability\\\": 77, \\\"supervision\\\": 78, \\\"contamination\\\": 45}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0050:noul", "group_id": "wf12:laboratory_batch:test:0050", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0a4a5287c9.\\nB means traceability >= 72.\\nD means contamination >= 28.\\nA means purity >= 19.\\nC means supervision >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 62, \\\"traceability\\\": 77, \\\"supervision\\\": 78, \\\"contamination\\\": 45}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0050:uncertain", "group_id": "wf12:laboratory_batch:test:0050", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0a4a5287c9.\\nB means traceability >= 72.\\nD means contamination >= 28.\\nA means purity >= 19.\\nC means supervision >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 62, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 78, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0050000000000000044, \"1\": 0.04500000000000004, \"2\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0051:choice", "group_id": "wf12:laboratory_batch:test:0051", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference bcd4930b8e.\\nB means traceability >= 48.\\nA means purity >= 48.\\nD means contamination >= 85.\\nC means supervision >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 86, \\\"supervision\\\": 45, \\\"contamination\\\": 28}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8df149\": \"Outcome PROCEED.\", \"candidate_70b0a0\": \"Outcome REVIEW.\", \"candidate_643bdd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8df149\": 1.0, \"candidate_70b0a0\": 0.0, \"candidate_643bdd\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0051:score", "group_id": "wf12:laboratory_batch:test:0051", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference bcd4930b8e.\\nB means traceability >= 48.\\nA means purity >= 48.\\nD means contamination >= 85.\\nC means supervision >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 86, \\\"supervision\\\": 45, \\\"contamination\\\": 28}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0051:noul", "group_id": "wf12:laboratory_batch:test:0051", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference bcd4930b8e.\\nB means traceability >= 48.\\nA means purity >= 48.\\nD means contamination >= 85.\\nC means supervision >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 97, \\\"traceability\\\": 86, \\\"supervision\\\": 45, \\\"contamination\\\": 28}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0051:uncertain", "group_id": "wf12:laboratory_batch:test:0051", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference bcd4930b8e.\\nB means traceability >= 48.\\nA means purity >= 48.\\nD means contamination >= 85.\\nC means supervision >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.8, \\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.93, \"true\": 0.06999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0052:choice", "group_id": "wf12:laboratory_batch:test:0052", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference bed85fe468.\\nB means traceability >= 61.\\nD means contamination >= 68.\\nA means purity >= 60.\\nC means supervision >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 79, \\\"supervision\\\": 31, \\\"contamination\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_021a46\": \"Outcome PROCEED.\", \"candidate_983503\": \"Outcome REVIEW.\", \"candidate_2028ef\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_021a46\": 0.0, \"candidate_983503\": 1.0, \"candidate_2028ef\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0052:score", "group_id": "wf12:laboratory_batch:test:0052", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference bed85fe468.\\nB means traceability >= 61.\\nD means contamination >= 68.\\nA means purity >= 60.\\nC means supervision >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 79, \\\"supervision\\\": 31, \\\"contamination\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0052:noul", "group_id": "wf12:laboratory_batch:test:0052", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference bed85fe468.\\nB means traceability >= 61.\\nD means contamination >= 68.\\nA means purity >= 60.\\nC means supervision >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 79, \\\"supervision\\\": 31, \\\"contamination\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0052:uncertain", "group_id": "wf12:laboratory_batch:test:0052", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference bed85fe468.\\nB means traceability >= 61.\\nD means contamination >= 68.\\nA means purity >= 60.\\nC means supervision >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.8, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.13}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0053:choice", "group_id": "wf12:laboratory_batch:test:0053", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d01cd8408e.\\nC means supervision >= 42.\\nD means contamination >= 66.\\nA means purity >= 69.\\nB means traceability >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 39, \\\"supervision\\\": 3, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d0bfde\": \"Outcome PROCEED.\", \"candidate_0565a8\": \"Outcome REVIEW.\", \"candidate_3804ca\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d0bfde\": 0.0, \"candidate_0565a8\": 0.0, \"candidate_3804ca\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0053:score", "group_id": "wf12:laboratory_batch:test:0053", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d01cd8408e.\\nC means supervision >= 42.\\nD means contamination >= 66.\\nA means purity >= 69.\\nB means traceability >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 39, \\\"supervision\\\": 3, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0053:noul", "group_id": "wf12:laboratory_batch:test:0053", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d01cd8408e.\\nC means supervision >= 42.\\nD means contamination >= 66.\\nA means purity >= 69.\\nB means traceability >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 39, \\\"supervision\\\": 3, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0053:uncertain", "group_id": "wf12:laboratory_batch:test:0053", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d01cd8408e.\\nC means supervision >= 42.\\nD means contamination >= 66.\\nA means purity >= 69.\\nB means traceability >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 39, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.9, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.995, \"true\": 0.004999999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0054:choice", "group_id": "wf12:laboratory_batch:test:0054", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a63269cb55.\\nD means contamination >= 72.\\nA means purity >= 67.\\nB means traceability >= 56.\\nC means supervision >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 56, \\\"supervision\\\": 95, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_02ca4d\": \"Outcome PROCEED.\", \"candidate_d3ec39\": \"Outcome REVIEW.\", \"candidate_1f8e6e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_02ca4d\": 1.0, \"candidate_d3ec39\": 0.0, \"candidate_1f8e6e\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0054:score", "group_id": "wf12:laboratory_batch:test:0054", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a63269cb55.\\nD means contamination >= 72.\\nA means purity >= 67.\\nB means traceability >= 56.\\nC means supervision >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 56, \\\"supervision\\\": 95, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0054:noul", "group_id": "wf12:laboratory_batch:test:0054", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a63269cb55.\\nD means contamination >= 72.\\nA means purity >= 67.\\nB means traceability >= 56.\\nC means supervision >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 56, \\\"supervision\\\": 95, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0054:uncertain", "group_id": "wf12:laboratory_batch:test:0054", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a63269cb55.\\nD means contamination >= 72.\\nA means purity >= 67.\\nB means traceability >= 56.\\nC means supervision >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.2, \\\"C\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_02ca4d\": \"Outcome PROCEED.\", \"candidate_d3ec39\": \"Outcome REVIEW.\", \"candidate_1f8e6e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_02ca4d\": 0.1412, \"candidate_d3ec39\": 0.5788000000000001, \"candidate_1f8e6e\": 0.28}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0055:choice", "group_id": "wf12:laboratory_batch:test:0055", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e61db6789.\\nA means purity >= 48.\\nC means supervision >= 65.\\nB means traceability >= 17.\\nD means contamination >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 37, \\\"traceability\\\": 12, \\\"supervision\\\": 72, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7b647e\": \"Outcome PROCEED.\", \"candidate_5bda2f\": \"Outcome REVIEW.\", \"candidate_2858ff\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7b647e\": 0.0, \"candidate_5bda2f\": 1.0, \"candidate_2858ff\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0055:score", "group_id": "wf12:laboratory_batch:test:0055", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e61db6789.\\nA means purity >= 48.\\nC means supervision >= 65.\\nB means traceability >= 17.\\nD means contamination >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 37, \\\"traceability\\\": 12, \\\"supervision\\\": 72, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0055:noul", "group_id": "wf12:laboratory_batch:test:0055", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e61db6789.\\nA means purity >= 48.\\nC means supervision >= 65.\\nB means traceability >= 17.\\nD means contamination >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 37, \\\"traceability\\\": 12, \\\"supervision\\\": 72, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0055:uncertain", "group_id": "wf12:laboratory_batch:test:0055", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e61db6789.\\nA means purity >= 48.\\nC means supervision >= 65.\\nB means traceability >= 17.\\nD means contamination >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.8, \\\"C\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7b647e\": \"Outcome PROCEED.\", \"candidate_5bda2f\": \"Outcome REVIEW.\", \"candidate_2858ff\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7b647e\": 0.720704, \"candidate_5bda2f\": 0.20929599999999995, \"candidate_2858ff\": 0.07}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0056:choice", "group_id": "wf12:laboratory_batch:test:0056", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8c9278f50c.\\nD means contamination >= 43.\\nA means purity >= 47.\\nC means supervision >= 18.\\nB means traceability >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 72, \\\"traceability\\\": 10, \\\"supervision\\\": 11, \\\"contamination\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d8dbcf\": \"Outcome PROCEED.\", \"candidate_118a44\": \"Outcome REVIEW.\", \"candidate_d2f5d0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d8dbcf\": 0.0, \"candidate_118a44\": 0.0, \"candidate_d2f5d0\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0056:score", "group_id": "wf12:laboratory_batch:test:0056", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8c9278f50c.\\nD means contamination >= 43.\\nA means purity >= 47.\\nC means supervision >= 18.\\nB means traceability >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 72, \\\"traceability\\\": 10, \\\"supervision\\\": 11, \\\"contamination\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0056:noul", "group_id": "wf12:laboratory_batch:test:0056", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8c9278f50c.\\nD means contamination >= 43.\\nA means purity >= 47.\\nC means supervision >= 18.\\nB means traceability >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 72, \\\"traceability\\\": 10, \\\"supervision\\\": 11, \\\"contamination\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0056:uncertain", "group_id": "wf12:laboratory_batch:test:0056", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8c9278f50c.\\nD means contamination >= 43.\\nA means purity >= 47.\\nC means supervision >= 18.\\nB means traceability >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 72, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 11, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d8dbcf\": \"Outcome PROCEED.\", \"candidate_118a44\": \"Outcome REVIEW.\", \"candidate_d2f5d0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d8dbcf\": 0.855, \"candidate_118a44\": 0.04500000000000004, \"candidate_d2f5d0\": 0.1}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0057:choice", "group_id": "wf12:laboratory_batch:test:0057", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4e2c9c2758.\\nC means supervision >= 16.\\nB means traceability >= 62.\\nD means contamination >= 56.\\nA means purity >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 86, \\\"supervision\\\": 11, \\\"contamination\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_768814\": \"Outcome PROCEED.\", \"candidate_9bc5ef\": \"Outcome REVIEW.\", \"candidate_539b06\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_768814\": 1.0, \"candidate_9bc5ef\": 0.0, \"candidate_539b06\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0057:score", "group_id": "wf12:laboratory_batch:test:0057", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4e2c9c2758.\\nC means supervision >= 16.\\nB means traceability >= 62.\\nD means contamination >= 56.\\nA means purity >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 86, \\\"supervision\\\": 11, \\\"contamination\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0057:noul", "group_id": "wf12:laboratory_batch:test:0057", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4e2c9c2758.\\nC means supervision >= 16.\\nB means traceability >= 62.\\nD means contamination >= 56.\\nA means purity >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 86, \\\"supervision\\\": 11, \\\"contamination\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0057:uncertain", "group_id": "wf12:laboratory_batch:test:0057", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4e2c9c2758.\\nC means supervision >= 16.\\nB means traceability >= 62.\\nD means contamination >= 56.\\nA means purity >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 11, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.42250000000000004, \"1\": 0.22749999999999998, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0058:choice", "group_id": "wf12:laboratory_batch:test:0058", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e2aa009129.\\nD means contamination >= 54.\\nC means supervision >= 54.\\nB means traceability >= 10.\\nA means purity >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 96, \\\"traceability\\\": 9, \\\"supervision\\\": 63, \\\"contamination\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_549dd3\": \"Outcome PROCEED.\", \"candidate_657e68\": \"Outcome REVIEW.\", \"candidate_33ba74\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_549dd3\": 0.0, \"candidate_657e68\": 1.0, \"candidate_33ba74\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0058:score", "group_id": "wf12:laboratory_batch:test:0058", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e2aa009129.\\nD means contamination >= 54.\\nC means supervision >= 54.\\nB means traceability >= 10.\\nA means purity >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 96, \\\"traceability\\\": 9, \\\"supervision\\\": 63, \\\"contamination\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0058:noul", "group_id": "wf12:laboratory_batch:test:0058", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e2aa009129.\\nD means contamination >= 54.\\nC means supervision >= 54.\\nB means traceability >= 10.\\nA means purity >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 96, \\\"traceability\\\": 9, \\\"supervision\\\": 63, \\\"contamination\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0058:uncertain", "group_id": "wf12:laboratory_batch:test:0058", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e2aa009129.\\nD means contamination >= 54.\\nC means supervision >= 54.\\nB means traceability >= 10.\\nA means purity >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 9, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 82}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.8, \"2\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0059:choice", "group_id": "wf12:laboratory_batch:test:0059", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6ff0ca8d56.\\nA means purity >= 73.\\nD means contamination >= 45.\\nC means supervision >= 49.\\nB means traceability >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 3, \\\"supervision\\\": 16, \\\"contamination\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ba7cf9\": \"Outcome PROCEED.\", \"candidate_0837cd\": \"Outcome REVIEW.\", \"candidate_f36487\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ba7cf9\": 0.0, \"candidate_0837cd\": 0.0, \"candidate_f36487\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0059:score", "group_id": "wf12:laboratory_batch:test:0059", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6ff0ca8d56.\\nA means purity >= 73.\\nD means contamination >= 45.\\nC means supervision >= 49.\\nB means traceability >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 3, \\\"supervision\\\": 16, \\\"contamination\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0059:noul", "group_id": "wf12:laboratory_batch:test:0059", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6ff0ca8d56.\\nA means purity >= 73.\\nD means contamination >= 45.\\nC means supervision >= 49.\\nB means traceability >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 3, \\\"supervision\\\": 16, \\\"contamination\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0059:uncertain", "group_id": "wf12:laboratory_batch:test:0059", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6ff0ca8d56.\\nA means purity >= 73.\\nD means contamination >= 45.\\nC means supervision >= 49.\\nB means traceability >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 68}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.05, \"2\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0060:choice", "group_id": "wf12:laboratory_batch:test:0060", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3fe3cb79eb.\\nC means supervision >= 44.\\nD means contamination >= 22.\\nA means purity >= 35.\\nB means traceability >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 42, \\\"traceability\\\": 0, \\\"supervision\\\": 88, \\\"contamination\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9858db\": \"Outcome PROCEED.\", \"candidate_245282\": \"Outcome REVIEW.\", \"candidate_28bd01\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9858db\": 1.0, \"candidate_245282\": 0.0, \"candidate_28bd01\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0060:score", "group_id": "wf12:laboratory_batch:test:0060", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3fe3cb79eb.\\nC means supervision >= 44.\\nD means contamination >= 22.\\nA means purity >= 35.\\nB means traceability >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 42, \\\"traceability\\\": 0, \\\"supervision\\\": 88, \\\"contamination\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0060:noul", "group_id": "wf12:laboratory_batch:test:0060", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3fe3cb79eb.\\nC means supervision >= 44.\\nD means contamination >= 22.\\nA means purity >= 35.\\nB means traceability >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 42, \\\"traceability\\\": 0, \\\"supervision\\\": 88, \\\"contamination\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0060:uncertain", "group_id": "wf12:laboratory_batch:test:0060", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3fe3cb79eb.\\nC means supervision >= 44.\\nD means contamination >= 22.\\nA means purity >= 35.\\nB means traceability >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.2, \\\"C\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9299999999999999, \"true\": 0.07}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0061:choice", "group_id": "wf12:laboratory_batch:test:0061", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0a90c644b1.\\nC means supervision >= 42.\\nB means traceability >= 48.\\nA means purity >= 69.\\nD means contamination >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 67, \\\"traceability\\\": 96, \\\"supervision\\\": 36, \\\"contamination\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d1288e\": \"Outcome PROCEED.\", \"candidate_e6bc11\": \"Outcome REVIEW.\", \"candidate_eb9301\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d1288e\": 0.0, \"candidate_e6bc11\": 1.0, \"candidate_eb9301\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0061:score", "group_id": "wf12:laboratory_batch:test:0061", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0a90c644b1.\\nC means supervision >= 42.\\nB means traceability >= 48.\\nA means purity >= 69.\\nD means contamination >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 67, \\\"traceability\\\": 96, \\\"supervision\\\": 36, \\\"contamination\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0061:noul", "group_id": "wf12:laboratory_batch:test:0061", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0a90c644b1.\\nC means supervision >= 42.\\nB means traceability >= 48.\\nA means purity >= 69.\\nD means contamination >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 67, \\\"traceability\\\": 96, \\\"supervision\\\": 36, \\\"contamination\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0061:uncertain", "group_id": "wf12:laboratory_batch:test:0061", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0a90c644b1.\\nC means supervision >= 42.\\nB means traceability >= 48.\\nA means purity >= 69.\\nD means contamination >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 67, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 36, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0062:choice", "group_id": "wf12:laboratory_batch:test:0062", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dfcace8ac1.\\nC means supervision >= 12.\\nD means contamination >= 76.\\nB means traceability >= 48.\\nA means purity >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 52, \\\"traceability\\\": 49, \\\"supervision\\\": 7, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b206d4\": \"Outcome PROCEED.\", \"candidate_4fc074\": \"Outcome REVIEW.\", \"candidate_799e09\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b206d4\": 0.0, \"candidate_4fc074\": 0.0, \"candidate_799e09\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0062:score", "group_id": "wf12:laboratory_batch:test:0062", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dfcace8ac1.\\nC means supervision >= 12.\\nD means contamination >= 76.\\nB means traceability >= 48.\\nA means purity >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 52, \\\"traceability\\\": 49, \\\"supervision\\\": 7, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0062:noul", "group_id": "wf12:laboratory_batch:test:0062", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dfcace8ac1.\\nC means supervision >= 12.\\nD means contamination >= 76.\\nB means traceability >= 48.\\nA means purity >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 52, \\\"traceability\\\": 49, \\\"supervision\\\": 7, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0062:uncertain", "group_id": "wf12:laboratory_batch:test:0062", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dfcace8ac1.\\nC means supervision >= 12.\\nD means contamination >= 76.\\nB means traceability >= 48.\\nA means purity >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 49, \\\"supervision\\\": 7, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09999999999999998, \"true\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0063:choice", "group_id": "wf12:laboratory_batch:test:0063", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 663f891fc4.\\nC means supervision >= 44.\\nB means traceability >= 81.\\nD means contamination >= 74.\\nA means purity >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 35, \\\"traceability\\\": 98, \\\"supervision\\\": 2, \\\"contamination\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_59c76e\": \"Outcome PROCEED.\", \"candidate_d7d6d4\": \"Outcome REVIEW.\", \"candidate_b97625\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_59c76e\": 1.0, \"candidate_d7d6d4\": 0.0, \"candidate_b97625\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0063:score", "group_id": "wf12:laboratory_batch:test:0063", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 663f891fc4.\\nC means supervision >= 44.\\nB means traceability >= 81.\\nD means contamination >= 74.\\nA means purity >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 35, \\\"traceability\\\": 98, \\\"supervision\\\": 2, \\\"contamination\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0063:noul", "group_id": "wf12:laboratory_batch:test:0063", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 663f891fc4.\\nC means supervision >= 44.\\nB means traceability >= 81.\\nD means contamination >= 74.\\nA means purity >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 35, \\\"traceability\\\": 98, \\\"supervision\\\": 2, \\\"contamination\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0063:uncertain", "group_id": "wf12:laboratory_batch:test:0063", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 663f891fc4.\\nC means supervision >= 44.\\nB means traceability >= 81.\\nD means contamination >= 74.\\nA means purity >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 35, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 2, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_59c76e\": \"Outcome PROCEED.\", \"candidate_d7d6d4\": \"Outcome REVIEW.\", \"candidate_b97625\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_59c76e\": 0.27999999999999997, \"candidate_d7d6d4\": 0.06999999999999998, \"candidate_b97625\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0064:choice", "group_id": "wf12:laboratory_batch:test:0064", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 63d5a38de2.\\nC means supervision >= 60.\\nD means contamination >= 20.\\nA means purity >= 79.\\nB means traceability >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 6, \\\"traceability\\\": 74, \\\"supervision\\\": 27, \\\"contamination\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_119f29\": \"Outcome PROCEED.\", \"candidate_157b2c\": \"Outcome REVIEW.\", \"candidate_e94ed4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_119f29\": 0.0, \"candidate_157b2c\": 1.0, \"candidate_e94ed4\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0064:score", "group_id": "wf12:laboratory_batch:test:0064", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 63d5a38de2.\\nC means supervision >= 60.\\nD means contamination >= 20.\\nA means purity >= 79.\\nB means traceability >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 6, \\\"traceability\\\": 74, \\\"supervision\\\": 27, \\\"contamination\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0064:noul", "group_id": "wf12:laboratory_batch:test:0064", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 63d5a38de2.\\nC means supervision >= 60.\\nD means contamination >= 20.\\nA means purity >= 79.\\nB means traceability >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 6, \\\"traceability\\\": 74, \\\"supervision\\\": 27, \\\"contamination\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0064:uncertain", "group_id": "wf12:laboratory_batch:test:0064", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 63d5a38de2.\\nC means supervision >= 60.\\nD means contamination >= 20.\\nA means purity >= 79.\\nB means traceability >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 6, \\\"traceability\\\": 74, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.2, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_119f29\": \"Outcome PROCEED.\", \"candidate_157b2c\": \"Outcome REVIEW.\", \"candidate_e94ed4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_119f29\": 0.27999999999999997, \"candidate_157b2c\": 0.59, \"candidate_e94ed4\": 0.13}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0065:choice", "group_id": "wf12:laboratory_batch:test:0065", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3d3c986e17.\\nA means purity >= 60.\\nD means contamination >= 38.\\nB means traceability >= 39.\\nC means supervision >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 51, \\\"traceability\\\": 2, \\\"supervision\\\": 36, \\\"contamination\\\": 66}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fde142\": \"Outcome PROCEED.\", \"candidate_7a7014\": \"Outcome REVIEW.\", \"candidate_b275c8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fde142\": 0.0, \"candidate_7a7014\": 0.0, \"candidate_b275c8\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0065:score", "group_id": "wf12:laboratory_batch:test:0065", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3d3c986e17.\\nA means purity >= 60.\\nD means contamination >= 38.\\nB means traceability >= 39.\\nC means supervision >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 51, \\\"traceability\\\": 2, \\\"supervision\\\": 36, \\\"contamination\\\": 66}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0065:noul", "group_id": "wf12:laboratory_batch:test:0065", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3d3c986e17.\\nA means purity >= 60.\\nD means contamination >= 38.\\nB means traceability >= 39.\\nC means supervision >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 51, \\\"traceability\\\": 2, \\\"supervision\\\": 36, \\\"contamination\\\": 66}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0065:uncertain", "group_id": "wf12:laboratory_batch:test:0065", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3d3c986e17.\\nA means purity >= 60.\\nD means contamination >= 38.\\nB means traceability >= 39.\\nC means supervision >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 51, \\\"traceability\\\": 2, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.05, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fde142\": \"Outcome PROCEED.\", \"candidate_7a7014\": \"Outcome REVIEW.\", \"candidate_b275c8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fde142\": 0.0025000000000000022, \"candidate_7a7014\": 0.09500000000000004, \"candidate_b275c8\": 0.9025}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0066:choice", "group_id": "wf12:laboratory_batch:test:0066", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f892f444ba.\\nA means purity >= 12.\\nC means supervision >= 56.\\nB means traceability >= 26.\\nD means contamination >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 90, \\\"supervision\\\": 1, \\\"contamination\\\": 19}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7feebc\": \"Outcome PROCEED.\", \"candidate_6340ae\": \"Outcome REVIEW.\", \"candidate_1596eb\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7feebc\": 1.0, \"candidate_6340ae\": 0.0, \"candidate_1596eb\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0066:score", "group_id": "wf12:laboratory_batch:test:0066", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f892f444ba.\\nA means purity >= 12.\\nC means supervision >= 56.\\nB means traceability >= 26.\\nD means contamination >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 90, \\\"supervision\\\": 1, \\\"contamination\\\": 19}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0066:noul", "group_id": "wf12:laboratory_batch:test:0066", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f892f444ba.\\nA means purity >= 12.\\nC means supervision >= 56.\\nB means traceability >= 26.\\nD means contamination >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 90, \\\"supervision\\\": 1, \\\"contamination\\\": 19}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0066:uncertain", "group_id": "wf12:laboratory_batch:test:0066", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f892f444ba.\\nA means purity >= 12.\\nC means supervision >= 56.\\nB means traceability >= 26.\\nD means contamination >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 90, \\\"supervision\\\": 1, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.06999999999999998, \"1\": 0.12999999999999998, \"2\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0067:choice", "group_id": "wf12:laboratory_batch:test:0067", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3c6f777e4e.\\nD means contamination >= 56.\\nB means traceability >= 83.\\nA means purity >= 83.\\nC means supervision >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 76, \\\"supervision\\\": 9, \\\"contamination\\\": 28}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_39b06c\": \"Outcome PROCEED.\", \"candidate_9ac979\": \"Outcome REVIEW.\", \"candidate_26a1e1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_39b06c\": 0.0, \"candidate_9ac979\": 1.0, \"candidate_26a1e1\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0067:score", "group_id": "wf12:laboratory_batch:test:0067", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3c6f777e4e.\\nD means contamination >= 56.\\nB means traceability >= 83.\\nA means purity >= 83.\\nC means supervision >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 76, \\\"supervision\\\": 9, \\\"contamination\\\": 28}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0067:noul", "group_id": "wf12:laboratory_batch:test:0067", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3c6f777e4e.\\nD means contamination >= 56.\\nB means traceability >= 83.\\nA means purity >= 83.\\nC means supervision >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": 76, \\\"supervision\\\": 9, \\\"contamination\\\": 28}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0067:uncertain", "group_id": "wf12:laboratory_batch:test:0067", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3c6f777e4e.\\nD means contamination >= 56.\\nB means traceability >= 83.\\nA means purity >= 83.\\nC means supervision >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 60, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 28}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.2, \"1\": 0.8, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0068:choice", "group_id": "wf12:laboratory_batch:test:0068", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7838ad1266.\\nB means traceability >= 89.\\nA means purity >= 90.\\nC means supervision >= 60.\\nD means contamination >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 94, \\\"supervision\\\": 16, \\\"contamination\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ddbc17\": \"Outcome PROCEED.\", \"candidate_fcd877\": \"Outcome REVIEW.\", \"candidate_450c76\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ddbc17\": 0.0, \"candidate_fcd877\": 0.0, \"candidate_450c76\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0068:score", "group_id": "wf12:laboratory_batch:test:0068", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7838ad1266.\\nB means traceability >= 89.\\nA means purity >= 90.\\nC means supervision >= 60.\\nD means contamination >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 94, \\\"supervision\\\": 16, \\\"contamination\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0068:noul", "group_id": "wf12:laboratory_batch:test:0068", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7838ad1266.\\nB means traceability >= 89.\\nA means purity >= 90.\\nC means supervision >= 60.\\nD means contamination >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 94, \\\"supervision\\\": 16, \\\"contamination\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0068:uncertain", "group_id": "wf12:laboratory_batch:test:0068", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7838ad1266.\\nB means traceability >= 89.\\nA means purity >= 90.\\nC means supervision >= 60.\\nD means contamination >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 94, \\\"supervision\\\": 16, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.045000000000000005, \"1\": 0.855, \"2\": 0.1}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0069:choice", "group_id": "wf12:laboratory_batch:test:0069", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5c23d250da.\\nA means purity >= 68.\\nB means traceability >= 12.\\nC means supervision >= 57.\\nD means contamination >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 90, \\\"traceability\\\": 3, \\\"supervision\\\": 75, \\\"contamination\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a45a02\": \"Outcome PROCEED.\", \"candidate_d558bb\": \"Outcome REVIEW.\", \"candidate_ab9cd6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a45a02\": 1.0, \"candidate_d558bb\": 0.0, \"candidate_ab9cd6\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0069:score", "group_id": "wf12:laboratory_batch:test:0069", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5c23d250da.\\nA means purity >= 68.\\nB means traceability >= 12.\\nC means supervision >= 57.\\nD means contamination >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 90, \\\"traceability\\\": 3, \\\"supervision\\\": 75, \\\"contamination\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0069:noul", "group_id": "wf12:laboratory_batch:test:0069", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5c23d250da.\\nA means purity >= 68.\\nB means traceability >= 12.\\nC means supervision >= 57.\\nD means contamination >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 90, \\\"traceability\\\": 3, \\\"supervision\\\": 75, \\\"contamination\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0069:uncertain", "group_id": "wf12:laboratory_batch:test:0069", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5c23d250da.\\nA means purity >= 68.\\nB means traceability >= 12.\\nC means supervision >= 57.\\nD means contamination >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.65, \\\"C\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.72, \"true\": 0.27999999999999997}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0070:choice", "group_id": "wf12:laboratory_batch:test:0070", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 70c9bff9b8.\\nA means purity >= 68.\\nB means traceability >= 69.\\nC means supervision >= 45.\\nD means contamination >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 19, \\\"traceability\\\": 81, \\\"supervision\\\": 94, \\\"contamination\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c69187\": \"Outcome PROCEED.\", \"candidate_857f9e\": \"Outcome REVIEW.\", \"candidate_eabff0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c69187\": 0.0, \"candidate_857f9e\": 1.0, \"candidate_eabff0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0070:score", "group_id": "wf12:laboratory_batch:test:0070", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 70c9bff9b8.\\nA means purity >= 68.\\nB means traceability >= 69.\\nC means supervision >= 45.\\nD means contamination >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 19, \\\"traceability\\\": 81, \\\"supervision\\\": 94, \\\"contamination\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0070:noul", "group_id": "wf12:laboratory_batch:test:0070", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 70c9bff9b8.\\nA means purity >= 68.\\nB means traceability >= 69.\\nC means supervision >= 45.\\nD means contamination >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 19, \\\"traceability\\\": 81, \\\"supervision\\\": 94, \\\"contamination\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0070:uncertain", "group_id": "wf12:laboratory_batch:test:0070", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 70c9bff9b8.\\nA means purity >= 68.\\nB means traceability >= 69.\\nC means supervision >= 45.\\nD means contamination >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65, \\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8775000000000001, \"true\": 0.12249999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0071:choice", "group_id": "wf12:laboratory_batch:test:0071", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7826753cc8.\\nC means supervision >= 49.\\nB means traceability >= 61.\\nD means contamination >= 22.\\nA means purity >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": 58, \\\"supervision\\\": 12, \\\"contamination\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0c0438\": \"Outcome PROCEED.\", \"candidate_3b671c\": \"Outcome REVIEW.\", \"candidate_9f377a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0c0438\": 0.0, \"candidate_3b671c\": 0.0, \"candidate_9f377a\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0071:score", "group_id": "wf12:laboratory_batch:test:0071", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7826753cc8.\\nC means supervision >= 49.\\nB means traceability >= 61.\\nD means contamination >= 22.\\nA means purity >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": 58, \\\"supervision\\\": 12, \\\"contamination\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0071:noul", "group_id": "wf12:laboratory_batch:test:0071", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7826753cc8.\\nC means supervision >= 49.\\nB means traceability >= 61.\\nD means contamination >= 22.\\nA means purity >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": 58, \\\"supervision\\\": 12, \\\"contamination\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0071:uncertain", "group_id": "wf12:laboratory_batch:test:0071", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7826753cc8.\\nC means supervision >= 49.\\nB means traceability >= 61.\\nD means contamination >= 22.\\nA means purity >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 5, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 12, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9, \"true\": 0.1}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0072:choice", "group_id": "wf12:laboratory_batch:test:0072", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 47fb200366.\\nD means contamination >= 73.\\nA means purity >= 18.\\nC means supervision >= 55.\\nB means traceability >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": 62, \\\"supervision\\\": 78, \\\"contamination\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d44bc7\": \"Outcome PROCEED.\", \"candidate_f8613e\": \"Outcome REVIEW.\", \"candidate_793838\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d44bc7\": 1.0, \"candidate_f8613e\": 0.0, \"candidate_793838\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0072:score", "group_id": "wf12:laboratory_batch:test:0072", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 47fb200366.\\nD means contamination >= 73.\\nA means purity >= 18.\\nC means supervision >= 55.\\nB means traceability >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": 62, \\\"supervision\\\": 78, \\\"contamination\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0072:noul", "group_id": "wf12:laboratory_batch:test:0072", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 47fb200366.\\nD means contamination >= 73.\\nA means purity >= 18.\\nC means supervision >= 55.\\nB means traceability >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": 62, \\\"supervision\\\": 78, \\\"contamination\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0072:uncertain", "group_id": "wf12:laboratory_batch:test:0072", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 47fb200366.\\nD means contamination >= 73.\\nA means purity >= 18.\\nC means supervision >= 55.\\nB means traceability >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 62, \\\"supervision\\\": 78, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d44bc7\": \"Outcome PROCEED.\", \"candidate_f8613e\": \"Outcome REVIEW.\", \"candidate_793838\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d44bc7\": 0.8, \"candidate_f8613e\": 0.2, \"candidate_793838\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0073:choice", "group_id": "wf12:laboratory_batch:test:0073", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d475cd15ca.\\nA means purity >= 85.\\nD means contamination >= 10.\\nC means supervision >= 45.\\nB means traceability >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 68, \\\"supervision\\\": 21, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b02d27\": \"Outcome PROCEED.\", \"candidate_181199\": \"Outcome REVIEW.\", \"candidate_4f0815\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b02d27\": 0.0, \"candidate_181199\": 1.0, \"candidate_4f0815\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0073:score", "group_id": "wf12:laboratory_batch:test:0073", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d475cd15ca.\\nA means purity >= 85.\\nD means contamination >= 10.\\nC means supervision >= 45.\\nB means traceability >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 68, \\\"supervision\\\": 21, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0073:noul", "group_id": "wf12:laboratory_batch:test:0073", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d475cd15ca.\\nA means purity >= 85.\\nD means contamination >= 10.\\nC means supervision >= 45.\\nB means traceability >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 68, \\\"supervision\\\": 21, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0073:uncertain", "group_id": "wf12:laboratory_batch:test:0073", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d475cd15ca.\\nA means purity >= 85.\\nD means contamination >= 10.\\nC means supervision >= 45.\\nB means traceability >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 68, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b02d27\": \"Outcome PROCEED.\", \"candidate_181199\": \"Outcome REVIEW.\", \"candidate_4f0815\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b02d27\": 0.22749999999999998, \"candidate_181199\": 0.545, \"candidate_4f0815\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0074:choice", "group_id": "wf12:laboratory_batch:test:0074", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4185b43474.\\nB means traceability >= 71.\\nC means supervision >= 37.\\nD means contamination >= 66.\\nA means purity >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 0, \\\"supervision\\\": 32, \\\"contamination\\\": 87}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_577b46\": \"Outcome PROCEED.\", \"candidate_118211\": \"Outcome REVIEW.\", \"candidate_13de2d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_577b46\": 0.0, \"candidate_118211\": 0.0, \"candidate_13de2d\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0074:score", "group_id": "wf12:laboratory_batch:test:0074", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4185b43474.\\nB means traceability >= 71.\\nC means supervision >= 37.\\nD means contamination >= 66.\\nA means purity >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 0, \\\"supervision\\\": 32, \\\"contamination\\\": 87}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0074:noul", "group_id": "wf12:laboratory_batch:test:0074", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4185b43474.\\nB means traceability >= 71.\\nC means supervision >= 37.\\nD means contamination >= 66.\\nA means purity >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 0, \\\"supervision\\\": 32, \\\"contamination\\\": 87}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0074:uncertain", "group_id": "wf12:laboratory_batch:test:0074", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4185b43474.\\nB means traceability >= 71.\\nC means supervision >= 37.\\nD means contamination >= 66.\\nA means purity >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.05, \\\"C\\\": 0.95, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_577b46\": \"Outcome PROCEED.\", \"candidate_118211\": \"Outcome REVIEW.\", \"candidate_13de2d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_577b46\": 0.09523749999999997, \"candidate_118211\": 0.8597625, \"candidate_13de2d\": 0.04500000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0075:choice", "group_id": "wf12:laboratory_batch:test:0075", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e366e8340.\\nB means traceability >= 62.\\nC means supervision >= 73.\\nA means purity >= 56.\\nD means contamination >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 80, \\\"traceability\\\": 67, \\\"supervision\\\": 53, \\\"contamination\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_07ba3a\": \"Outcome PROCEED.\", \"candidate_2651b4\": \"Outcome REVIEW.\", \"candidate_41b441\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_07ba3a\": 1.0, \"candidate_2651b4\": 0.0, \"candidate_41b441\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0075:score", "group_id": "wf12:laboratory_batch:test:0075", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e366e8340.\\nB means traceability >= 62.\\nC means supervision >= 73.\\nA means purity >= 56.\\nD means contamination >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 80, \\\"traceability\\\": 67, \\\"supervision\\\": 53, \\\"contamination\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0075:noul", "group_id": "wf12:laboratory_batch:test:0075", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e366e8340.\\nB means traceability >= 62.\\nC means supervision >= 73.\\nA means purity >= 56.\\nD means contamination >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 80, \\\"traceability\\\": 67, \\\"supervision\\\": 53, \\\"contamination\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0075:uncertain", "group_id": "wf12:laboratory_batch:test:0075", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3e366e8340.\\nB means traceability >= 62.\\nC means supervision >= 73.\\nA means purity >= 56.\\nD means contamination >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 80, \\\"traceability\\\": 67, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.2, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.13, \"2\": 0.52}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0076:choice", "group_id": "wf12:laboratory_batch:test:0076", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 53563a1492.\\nA means purity >= 31.\\nC means supervision >= 70.\\nB means traceability >= 14.\\nD means contamination >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 53, \\\"traceability\\\": 0, \\\"supervision\\\": 66, \\\"contamination\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_db11e7\": \"Outcome PROCEED.\", \"candidate_caa6e0\": \"Outcome REVIEW.\", \"candidate_dcea5a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_db11e7\": 0.0, \"candidate_caa6e0\": 1.0, \"candidate_dcea5a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0076:score", "group_id": "wf12:laboratory_batch:test:0076", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 53563a1492.\\nA means purity >= 31.\\nC means supervision >= 70.\\nB means traceability >= 14.\\nD means contamination >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 53, \\\"traceability\\\": 0, \\\"supervision\\\": 66, \\\"contamination\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0076:noul", "group_id": "wf12:laboratory_batch:test:0076", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 53563a1492.\\nA means purity >= 31.\\nC means supervision >= 70.\\nB means traceability >= 14.\\nD means contamination >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 53, \\\"traceability\\\": 0, \\\"supervision\\\": 66, \\\"contamination\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0076:uncertain", "group_id": "wf12:laboratory_batch:test:0076", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 53563a1492.\\nA means purity >= 31.\\nC means supervision >= 70.\\nB means traceability >= 14.\\nD means contamination >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 53, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 50}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8699999999999999, \"1\": 0.12999999999999998, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0077:choice", "group_id": "wf12:laboratory_batch:test:0077", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 82ac6d948e.\\nC means supervision >= 58.\\nD means contamination >= 77.\\nA means purity >= 36.\\nB means traceability >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 3, \\\"supervision\\\": 34, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bb6b2e\": \"Outcome PROCEED.\", \"candidate_efff18\": \"Outcome REVIEW.\", \"candidate_e02463\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bb6b2e\": 0.0, \"candidate_efff18\": 0.0, \"candidate_e02463\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0077:score", "group_id": "wf12:laboratory_batch:test:0077", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 82ac6d948e.\\nC means supervision >= 58.\\nD means contamination >= 77.\\nA means purity >= 36.\\nB means traceability >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 3, \\\"supervision\\\": 34, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0077:noul", "group_id": "wf12:laboratory_batch:test:0077", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 82ac6d948e.\\nC means supervision >= 58.\\nD means contamination >= 77.\\nA means purity >= 36.\\nB means traceability >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 3, \\\"supervision\\\": 34, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0077:uncertain", "group_id": "wf12:laboratory_batch:test:0077", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 82ac6d948e.\\nC means supervision >= 58.\\nD means contamination >= 77.\\nA means purity >= 36.\\nB means traceability >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 3, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.05, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0025000000000000022, \"1\": 0.09500000000000004, \"2\": 0.9025}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0078:choice", "group_id": "wf12:laboratory_batch:test:0078", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 596766805d.\\nC means supervision >= 49.\\nD means contamination >= 43.\\nB means traceability >= 12.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 74, \\\"supervision\\\": 83, \\\"contamination\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_14f2d6\": \"Outcome PROCEED.\", \"candidate_b6c9d9\": \"Outcome REVIEW.\", \"candidate_173d51\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_14f2d6\": 1.0, \"candidate_b6c9d9\": 0.0, \"candidate_173d51\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0078:score", "group_id": "wf12:laboratory_batch:test:0078", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 596766805d.\\nC means supervision >= 49.\\nD means contamination >= 43.\\nB means traceability >= 12.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 74, \\\"supervision\\\": 83, \\\"contamination\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0078:noul", "group_id": "wf12:laboratory_batch:test:0078", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 596766805d.\\nC means supervision >= 49.\\nD means contamination >= 43.\\nB means traceability >= 12.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 95, \\\"traceability\\\": 74, \\\"supervision\\\": 83, \\\"contamination\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0078:uncertain", "group_id": "wf12:laboratory_batch:test:0078", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 596766805d.\\nC means supervision >= 49.\\nD means contamination >= 43.\\nB means traceability >= 12.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.8, \\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.93, \"true\": 0.06999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0079:choice", "group_id": "wf12:laboratory_batch:test:0079", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference cc0172ce28.\\nB means traceability >= 75.\\nA means purity >= 18.\\nD means contamination >= 68.\\nC means supervision >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 22, \\\"traceability\\\": 10, \\\"supervision\\\": 19, \\\"contamination\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_873781\": \"Outcome PROCEED.\", \"candidate_08c026\": \"Outcome REVIEW.\", \"candidate_42bade\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_873781\": 0.0, \"candidate_08c026\": 1.0, \"candidate_42bade\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0079:score", "group_id": "wf12:laboratory_batch:test:0079", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference cc0172ce28.\\nB means traceability >= 75.\\nA means purity >= 18.\\nD means contamination >= 68.\\nC means supervision >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 22, \\\"traceability\\\": 10, \\\"supervision\\\": 19, \\\"contamination\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0079:noul", "group_id": "wf12:laboratory_batch:test:0079", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference cc0172ce28.\\nB means traceability >= 75.\\nA means purity >= 18.\\nD means contamination >= 68.\\nC means supervision >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 22, \\\"traceability\\\": 10, \\\"supervision\\\": 19, \\\"contamination\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0079:uncertain", "group_id": "wf12:laboratory_batch:test:0079", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference cc0172ce28.\\nB means traceability >= 75.\\nA means purity >= 18.\\nD means contamination >= 68.\\nC means supervision >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65, \\\"C\\\": 0.35, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.48, \"true\": 0.52}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0080:choice", "group_id": "wf12:laboratory_batch:test:0080", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fa50090887.\\nD means contamination >= 63.\\nB means traceability >= 77.\\nA means purity >= 49.\\nC means supervision >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 83, \\\"supervision\\\": 27, \\\"contamination\\\": 97}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5077f6\": \"Outcome PROCEED.\", \"candidate_2b73c7\": \"Outcome REVIEW.\", \"candidate_6fe72b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5077f6\": 0.0, \"candidate_2b73c7\": 0.0, \"candidate_6fe72b\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0080:score", "group_id": "wf12:laboratory_batch:test:0080", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fa50090887.\\nD means contamination >= 63.\\nB means traceability >= 77.\\nA means purity >= 49.\\nC means supervision >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 83, \\\"supervision\\\": 27, \\\"contamination\\\": 97}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0080:noul", "group_id": "wf12:laboratory_batch:test:0080", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fa50090887.\\nD means contamination >= 63.\\nB means traceability >= 77.\\nA means purity >= 49.\\nC means supervision >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 83, \\\"supervision\\\": 27, \\\"contamination\\\": 97}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0080:uncertain", "group_id": "wf12:laboratory_batch:test:0080", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference fa50090887.\\nD means contamination >= 63.\\nB means traceability >= 77.\\nA means purity >= 49.\\nC means supervision >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 83, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 97}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09999999999999998, \"true\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0081:choice", "group_id": "wf12:laboratory_batch:test:0081", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference aa6d180125.\\nB means traceability >= 43.\\nD means contamination >= 55.\\nC means supervision >= 29.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 31, \\\"supervision\\\": 58, \\\"contamination\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_00195b\": \"Outcome PROCEED.\", \"candidate_093ae0\": \"Outcome REVIEW.\", \"candidate_8abd62\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_00195b\": 1.0, \"candidate_093ae0\": 0.0, \"candidate_8abd62\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0081:score", "group_id": "wf12:laboratory_batch:test:0081", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference aa6d180125.\\nB means traceability >= 43.\\nD means contamination >= 55.\\nC means supervision >= 29.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 31, \\\"supervision\\\": 58, \\\"contamination\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0081:noul", "group_id": "wf12:laboratory_batch:test:0081", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference aa6d180125.\\nB means traceability >= 43.\\nD means contamination >= 55.\\nC means supervision >= 29.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": 31, \\\"supervision\\\": 58, \\\"contamination\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0081:uncertain", "group_id": "wf12:laboratory_batch:test:0081", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference aa6d180125.\\nB means traceability >= 43.\\nD means contamination >= 55.\\nC means supervision >= 29.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 3, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 31}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_00195b\": \"Outcome PROCEED.\", \"candidate_093ae0\": \"Outcome REVIEW.\", \"candidate_8abd62\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_00195b\": 0.8000000000000002, \"candidate_093ae0\": 0.19999999999999996, \"candidate_8abd62\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0082:choice", "group_id": "wf12:laboratory_batch:test:0082", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e694e4ef3f.\\nD means contamination >= 81.\\nC means supervision >= 70.\\nB means traceability >= 61.\\nA means purity >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 34, \\\"traceability\\\": 96, \\\"supervision\\\": 85, \\\"contamination\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4a1f0e\": \"Outcome PROCEED.\", \"candidate_d4f2d5\": \"Outcome REVIEW.\", \"candidate_c9755a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4a1f0e\": 0.0, \"candidate_d4f2d5\": 1.0, \"candidate_c9755a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0082:score", "group_id": "wf12:laboratory_batch:test:0082", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e694e4ef3f.\\nD means contamination >= 81.\\nC means supervision >= 70.\\nB means traceability >= 61.\\nA means purity >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 34, \\\"traceability\\\": 96, \\\"supervision\\\": 85, \\\"contamination\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0082:noul", "group_id": "wf12:laboratory_batch:test:0082", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e694e4ef3f.\\nD means contamination >= 81.\\nC means supervision >= 70.\\nB means traceability >= 61.\\nA means purity >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 34, \\\"traceability\\\": 96, \\\"supervision\\\": 85, \\\"contamination\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0082:uncertain", "group_id": "wf12:laboratory_batch:test:0082", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e694e4ef3f.\\nD means contamination >= 81.\\nC means supervision >= 70.\\nB means traceability >= 61.\\nA means purity >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 34, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 85, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4a1f0e\": \"Outcome PROCEED.\", \"candidate_d4f2d5\": \"Outcome REVIEW.\", \"candidate_c9755a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4a1f0e\": 0.65, \"candidate_d4f2d5\": 0.35, \"candidate_c9755a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0083:choice", "group_id": "wf12:laboratory_batch:test:0083", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b506c4a0bf.\\nA means purity >= 31.\\nB means traceability >= 79.\\nD means contamination >= 78.\\nC means supervision >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 24, \\\"traceability\\\": 100, \\\"supervision\\\": 48, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c1fcc5\": \"Outcome PROCEED.\", \"candidate_4a6dde\": \"Outcome REVIEW.\", \"candidate_4a8044\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c1fcc5\": 0.0, \"candidate_4a6dde\": 0.0, \"candidate_4a8044\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0083:score", "group_id": "wf12:laboratory_batch:test:0083", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b506c4a0bf.\\nA means purity >= 31.\\nB means traceability >= 79.\\nD means contamination >= 78.\\nC means supervision >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 24, \\\"traceability\\\": 100, \\\"supervision\\\": 48, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0083:noul", "group_id": "wf12:laboratory_batch:test:0083", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b506c4a0bf.\\nA means purity >= 31.\\nB means traceability >= 79.\\nD means contamination >= 78.\\nC means supervision >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 24, \\\"traceability\\\": 100, \\\"supervision\\\": 48, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0083:uncertain", "group_id": "wf12:laboratory_batch:test:0083", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b506c4a0bf.\\nA means purity >= 31.\\nB means traceability >= 79.\\nD means contamination >= 78.\\nC means supervision >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 100, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 89}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"C\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c1fcc5\": \"Outcome PROCEED.\", \"candidate_4a6dde\": \"Outcome REVIEW.\", \"candidate_4a8044\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c1fcc5\": 0.0, \"candidate_4a6dde\": 0.9, \"candidate_4a8044\": 0.09999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0084:choice", "group_id": "wf12:laboratory_batch:test:0084", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1ff482c020.\\nB means traceability >= 25.\\nC means supervision >= 73.\\nA means purity >= 17.\\nD means contamination >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 14, \\\"traceability\\\": 11, \\\"supervision\\\": 81, \\\"contamination\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7aeaeb\": \"Outcome PROCEED.\", \"candidate_7eea9e\": \"Outcome REVIEW.\", \"candidate_7fcada\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7aeaeb\": 1.0, \"candidate_7eea9e\": 0.0, \"candidate_7fcada\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0084:score", "group_id": "wf12:laboratory_batch:test:0084", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1ff482c020.\\nB means traceability >= 25.\\nC means supervision >= 73.\\nA means purity >= 17.\\nD means contamination >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 14, \\\"traceability\\\": 11, \\\"supervision\\\": 81, \\\"contamination\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0084:noul", "group_id": "wf12:laboratory_batch:test:0084", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1ff482c020.\\nB means traceability >= 25.\\nC means supervision >= 73.\\nA means purity >= 17.\\nD means contamination >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 14, \\\"traceability\\\": 11, \\\"supervision\\\": 81, \\\"contamination\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0084:uncertain", "group_id": "wf12:laboratory_batch:test:0084", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1ff482c020.\\nB means traceability >= 25.\\nC means supervision >= 73.\\nA means purity >= 17.\\nD means contamination >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 14, \\\"traceability\\\": 11, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.13, \"1\": 0.59, \"2\": 0.27999999999999997}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0085:choice", "group_id": "wf12:laboratory_batch:test:0085", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f87d94afd8.\\nB means traceability >= 82.\\nD means contamination >= 29.\\nA means purity >= 88.\\nC means supervision >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 23, \\\"traceability\\\": 86, \\\"supervision\\\": 83, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c62d4a\": \"Outcome PROCEED.\", \"candidate_3fc93a\": \"Outcome REVIEW.\", \"candidate_87ad4f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c62d4a\": 0.0, \"candidate_3fc93a\": 1.0, \"candidate_87ad4f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0085:score", "group_id": "wf12:laboratory_batch:test:0085", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f87d94afd8.\\nB means traceability >= 82.\\nD means contamination >= 29.\\nA means purity >= 88.\\nC means supervision >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 23, \\\"traceability\\\": 86, \\\"supervision\\\": 83, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0085:noul", "group_id": "wf12:laboratory_batch:test:0085", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f87d94afd8.\\nB means traceability >= 82.\\nD means contamination >= 29.\\nA means purity >= 88.\\nC means supervision >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 23, \\\"traceability\\\": 86, \\\"supervision\\\": 83, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0085:uncertain", "group_id": "wf12:laboratory_batch:test:0085", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference f87d94afd8.\\nB means traceability >= 82.\\nD means contamination >= 29.\\nA means purity >= 88.\\nC means supervision >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 23, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 99}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.65, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0086:choice", "group_id": "wf12:laboratory_batch:test:0086", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4db5401b89.\\nC means supervision >= 61.\\nB means traceability >= 10.\\nD means contamination >= 82.\\nA means purity >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 66, \\\"traceability\\\": 89, \\\"supervision\\\": 6, \\\"contamination\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_34c940\": \"Outcome PROCEED.\", \"candidate_b2e54d\": \"Outcome REVIEW.\", \"candidate_271005\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_34c940\": 0.0, \"candidate_b2e54d\": 0.0, \"candidate_271005\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0086:score", "group_id": "wf12:laboratory_batch:test:0086", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4db5401b89.\\nC means supervision >= 61.\\nB means traceability >= 10.\\nD means contamination >= 82.\\nA means purity >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 66, \\\"traceability\\\": 89, \\\"supervision\\\": 6, \\\"contamination\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0086:noul", "group_id": "wf12:laboratory_batch:test:0086", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4db5401b89.\\nC means supervision >= 61.\\nB means traceability >= 10.\\nD means contamination >= 82.\\nA means purity >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 66, \\\"traceability\\\": 89, \\\"supervision\\\": 6, \\\"contamination\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0086:uncertain", "group_id": "wf12:laboratory_batch:test:0086", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4db5401b89.\\nC means supervision >= 61.\\nB means traceability >= 10.\\nD means contamination >= 82.\\nA means purity >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"B\\\": 0.9, \\\"C\\\": 0.1, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.7875500000000002, \"1\": 0.16745, \"2\": 0.045}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0087:choice", "group_id": "wf12:laboratory_batch:test:0087", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7ce7fdddfb.\\nA means purity >= 77.\\nD means contamination >= 19.\\nB means traceability >= 36.\\nC means supervision >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 40, \\\"supervision\\\": 42, \\\"contamination\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fb65cb\": \"Outcome PROCEED.\", \"candidate_29fb87\": \"Outcome REVIEW.\", \"candidate_28899b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fb65cb\": 1.0, \"candidate_29fb87\": 0.0, \"candidate_28899b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0087:score", "group_id": "wf12:laboratory_batch:test:0087", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7ce7fdddfb.\\nA means purity >= 77.\\nD means contamination >= 19.\\nB means traceability >= 36.\\nC means supervision >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 40, \\\"supervision\\\": 42, \\\"contamination\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0087:noul", "group_id": "wf12:laboratory_batch:test:0087", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7ce7fdddfb.\\nA means purity >= 77.\\nD means contamination >= 19.\\nB means traceability >= 36.\\nC means supervision >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 36, \\\"traceability\\\": 40, \\\"supervision\\\": 42, \\\"contamination\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0087:uncertain", "group_id": "wf12:laboratory_batch:test:0087", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7ce7fdddfb.\\nA means purity >= 77.\\nD means contamination >= 19.\\nB means traceability >= 36.\\nC means supervision >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.8, \\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.93, \"true\": 0.06999999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0088:choice", "group_id": "wf12:laboratory_batch:test:0088", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference eaedb13923.\\nD means contamination >= 38.\\nC means supervision >= 78.\\nB means traceability >= 84.\\nA means purity >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": 98, \\\"supervision\\\": 81, \\\"contamination\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d44f03\": \"Outcome PROCEED.\", \"candidate_36245d\": \"Outcome REVIEW.\", \"candidate_983a1b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d44f03\": 0.0, \"candidate_36245d\": 1.0, \"candidate_983a1b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0088:score", "group_id": "wf12:laboratory_batch:test:0088", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference eaedb13923.\\nD means contamination >= 38.\\nC means supervision >= 78.\\nB means traceability >= 84.\\nA means purity >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": 98, \\\"supervision\\\": 81, \\\"contamination\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0088:noul", "group_id": "wf12:laboratory_batch:test:0088", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference eaedb13923.\\nD means contamination >= 38.\\nC means supervision >= 78.\\nB means traceability >= 84.\\nA means purity >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 8, \\\"traceability\\\": 98, \\\"supervision\\\": 81, \\\"contamination\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0088:uncertain", "group_id": "wf12:laboratory_batch:test:0088", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference eaedb13923.\\nD means contamination >= 38.\\nC means supervision >= 78.\\nB means traceability >= 84.\\nA means purity >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.8, \\\"C\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.12999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0089:choice", "group_id": "wf12:laboratory_batch:test:0089", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e6e01991c.\\nA means purity >= 86.\\nD means contamination >= 40.\\nB means traceability >= 42.\\nC means supervision >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 26, \\\"traceability\\\": 9, \\\"supervision\\\": 38, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_89dc34\": \"Outcome PROCEED.\", \"candidate_d55859\": \"Outcome REVIEW.\", \"candidate_e13b07\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_89dc34\": 0.0, \"candidate_d55859\": 0.0, \"candidate_e13b07\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0089:score", "group_id": "wf12:laboratory_batch:test:0089", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e6e01991c.\\nA means purity >= 86.\\nD means contamination >= 40.\\nB means traceability >= 42.\\nC means supervision >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 26, \\\"traceability\\\": 9, \\\"supervision\\\": 38, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0089:noul", "group_id": "wf12:laboratory_batch:test:0089", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e6e01991c.\\nA means purity >= 86.\\nD means contamination >= 40.\\nB means traceability >= 42.\\nC means supervision >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 26, \\\"traceability\\\": 9, \\\"supervision\\\": 38, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0089:uncertain", "group_id": "wf12:laboratory_batch:test:0089", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e6e01991c.\\nA means purity >= 86.\\nD means contamination >= 40.\\nB means traceability >= 42.\\nC means supervision >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 26, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 67}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"C\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9, \"true\": 0.09999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0090:choice", "group_id": "wf12:laboratory_batch:test:0090", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7d66bb0163.\\nC means supervision >= 87.\\nB means traceability >= 32.\\nA means purity >= 22.\\nD means contamination >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 25, \\\"traceability\\\": 77, \\\"supervision\\\": 52, \\\"contamination\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3481fe\": \"Outcome PROCEED.\", \"candidate_20f88f\": \"Outcome REVIEW.\", \"candidate_887678\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3481fe\": 1.0, \"candidate_20f88f\": 0.0, \"candidate_887678\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0090:score", "group_id": "wf12:laboratory_batch:test:0090", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7d66bb0163.\\nC means supervision >= 87.\\nB means traceability >= 32.\\nA means purity >= 22.\\nD means contamination >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 25, \\\"traceability\\\": 77, \\\"supervision\\\": 52, \\\"contamination\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0090:noul", "group_id": "wf12:laboratory_batch:test:0090", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7d66bb0163.\\nC means supervision >= 87.\\nB means traceability >= 32.\\nA means purity >= 22.\\nD means contamination >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 25, \\\"traceability\\\": 77, \\\"supervision\\\": 52, \\\"contamination\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0090:uncertain", "group_id": "wf12:laboratory_batch:test:0090", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 7d66bb0163.\\nC means supervision >= 87.\\nB means traceability >= 32.\\nA means purity >= 22.\\nD means contamination >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 25, \\\"traceability\\\": 77, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3481fe\": \"Outcome PROCEED.\", \"candidate_20f88f\": \"Outcome REVIEW.\", \"candidate_887678\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3481fe\": 0.65, \"candidate_20f88f\": 0.12249999999999998, \"candidate_887678\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0091:choice", "group_id": "wf12:laboratory_batch:test:0091", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 673fba7450.\\nB means traceability >= 84.\\nD means contamination >= 90.\\nA means purity >= 85.\\nC means supervision >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 54, \\\"supervision\\\": 79, \\\"contamination\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9a685b\": \"Outcome PROCEED.\", \"candidate_5ff5c3\": \"Outcome REVIEW.\", \"candidate_9277d3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9a685b\": 0.0, \"candidate_5ff5c3\": 1.0, \"candidate_9277d3\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0091:score", "group_id": "wf12:laboratory_batch:test:0091", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 673fba7450.\\nB means traceability >= 84.\\nD means contamination >= 90.\\nA means purity >= 85.\\nC means supervision >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 54, \\\"supervision\\\": 79, \\\"contamination\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0091:noul", "group_id": "wf12:laboratory_batch:test:0091", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 673fba7450.\\nB means traceability >= 84.\\nD means contamination >= 90.\\nA means purity >= 85.\\nC means supervision >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 54, \\\"supervision\\\": 79, \\\"contamination\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0091:uncertain", "group_id": "wf12:laboratory_batch:test:0091", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 673fba7450.\\nB means traceability >= 84.\\nD means contamination >= 90.\\nA means purity >= 85.\\nC means supervision >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.2, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9a685b\": \"Outcome PROCEED.\", \"candidate_5ff5c3\": \"Outcome REVIEW.\", \"candidate_9277d3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9a685b\": 0.31639999999999996, \"candidate_5ff5c3\": 0.5536, \"candidate_9277d3\": 0.13}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0092:choice", "group_id": "wf12:laboratory_batch:test:0092", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 67e97f31bb.\\nB means traceability >= 25.\\nC means supervision >= 75.\\nD means contamination >= 58.\\nA means purity >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 39, \\\"traceability\\\": 65, \\\"supervision\\\": 28, \\\"contamination\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dfe3d2\": \"Outcome PROCEED.\", \"candidate_020389\": \"Outcome REVIEW.\", \"candidate_de0697\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dfe3d2\": 0.0, \"candidate_020389\": 0.0, \"candidate_de0697\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0092:score", "group_id": "wf12:laboratory_batch:test:0092", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 67e97f31bb.\\nB means traceability >= 25.\\nC means supervision >= 75.\\nD means contamination >= 58.\\nA means purity >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 39, \\\"traceability\\\": 65, \\\"supervision\\\": 28, \\\"contamination\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0092:noul", "group_id": "wf12:laboratory_batch:test:0092", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 67e97f31bb.\\nB means traceability >= 25.\\nC means supervision >= 75.\\nD means contamination >= 58.\\nA means purity >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 39, \\\"traceability\\\": 65, \\\"supervision\\\": 28, \\\"contamination\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0092:uncertain", "group_id": "wf12:laboratory_batch:test:0092", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 67e97f31bb.\\nB means traceability >= 25.\\nC means supervision >= 75.\\nD means contamination >= 58.\\nA means purity >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 39, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 28, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dfe3d2\": \"Outcome PROCEED.\", \"candidate_020389\": \"Outcome REVIEW.\", \"candidate_de0697\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dfe3d2\": 0.0, \"candidate_020389\": 0.95, \"candidate_de0697\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0093:choice", "group_id": "wf12:laboratory_batch:test:0093", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 809849adb8.\\nD means contamination >= 37.\\nA means purity >= 83.\\nC means supervision >= 65.\\nB means traceability >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 88, \\\"traceability\\\": 32, \\\"supervision\\\": 74, \\\"contamination\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4f5e63\": \"Outcome PROCEED.\", \"candidate_681382\": \"Outcome REVIEW.\", \"candidate_1f9d72\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4f5e63\": 1.0, \"candidate_681382\": 0.0, \"candidate_1f9d72\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0093:score", "group_id": "wf12:laboratory_batch:test:0093", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 809849adb8.\\nD means contamination >= 37.\\nA means purity >= 83.\\nC means supervision >= 65.\\nB means traceability >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 88, \\\"traceability\\\": 32, \\\"supervision\\\": 74, \\\"contamination\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0093:noul", "group_id": "wf12:laboratory_batch:test:0093", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 809849adb8.\\nD means contamination >= 37.\\nA means purity >= 83.\\nC means supervision >= 65.\\nB means traceability >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 88, \\\"traceability\\\": 32, \\\"supervision\\\": 74, \\\"contamination\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0093:uncertain", "group_id": "wf12:laboratory_batch:test:0093", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 809849adb8.\\nD means contamination >= 37.\\nA means purity >= 83.\\nC means supervision >= 65.\\nB means traceability >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 88, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 3}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.5774999999999999, \"1\": 0.42250000000000004, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0094:choice", "group_id": "wf12:laboratory_batch:test:0094", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 769f3c301f.\\nB means traceability >= 38.\\nA means purity >= 74.\\nD means contamination >= 22.\\nC means supervision >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 92, \\\"traceability\\\": 6, \\\"supervision\\\": 57, \\\"contamination\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_74851a\": \"Outcome PROCEED.\", \"candidate_092902\": \"Outcome REVIEW.\", \"candidate_04b0b0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_74851a\": 0.0, \"candidate_092902\": 1.0, \"candidate_04b0b0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0094:score", "group_id": "wf12:laboratory_batch:test:0094", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 769f3c301f.\\nB means traceability >= 38.\\nA means purity >= 74.\\nD means contamination >= 22.\\nC means supervision >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 92, \\\"traceability\\\": 6, \\\"supervision\\\": 57, \\\"contamination\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0094:noul", "group_id": "wf12:laboratory_batch:test:0094", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 769f3c301f.\\nB means traceability >= 38.\\nA means purity >= 74.\\nD means contamination >= 22.\\nC means supervision >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 92, \\\"traceability\\\": 6, \\\"supervision\\\": 57, \\\"contamination\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0094:uncertain", "group_id": "wf12:laboratory_batch:test:0094", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 769f3c301f.\\nB means traceability >= 38.\\nA means purity >= 74.\\nD means contamination >= 22.\\nC means supervision >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 92, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 7}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.5774999999999999, \"1\": 0.42250000000000004, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0095:choice", "group_id": "wf12:laboratory_batch:test:0095", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dd92173924.\\nA means purity >= 67.\\nD means contamination >= 79.\\nB means traceability >= 54.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 80, \\\"traceability\\\": 98, \\\"supervision\\\": 44, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e7deb2\": \"Outcome PROCEED.\", \"candidate_03b860\": \"Outcome REVIEW.\", \"candidate_8d7c8c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e7deb2\": 0.0, \"candidate_03b860\": 0.0, \"candidate_8d7c8c\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0095:score", "group_id": "wf12:laboratory_batch:test:0095", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dd92173924.\\nA means purity >= 67.\\nD means contamination >= 79.\\nB means traceability >= 54.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 80, \\\"traceability\\\": 98, \\\"supervision\\\": 44, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0095:noul", "group_id": "wf12:laboratory_batch:test:0095", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dd92173924.\\nA means purity >= 67.\\nD means contamination >= 79.\\nB means traceability >= 54.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 80, \\\"traceability\\\": 98, \\\"supervision\\\": 44, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0095:uncertain", "group_id": "wf12:laboratory_batch:test:0095", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference dd92173924.\\nA means purity >= 67.\\nD means contamination >= 79.\\nB means traceability >= 54.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 80, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 44, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9025, \"1\": 0.04750000000000004, \"2\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0096:choice", "group_id": "wf12:laboratory_batch:test:0096", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d8618811da.\\nC means supervision >= 41.\\nA means purity >= 71.\\nD means contamination >= 61.\\nB means traceability >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 72, \\\"traceability\\\": 30, \\\"supervision\\\": 59, \\\"contamination\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_552d34\": \"Outcome PROCEED.\", \"candidate_a0488d\": \"Outcome REVIEW.\", \"candidate_6976c5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_552d34\": 1.0, \"candidate_a0488d\": 0.0, \"candidate_6976c5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0096:score", "group_id": "wf12:laboratory_batch:test:0096", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d8618811da.\\nC means supervision >= 41.\\nA means purity >= 71.\\nD means contamination >= 61.\\nB means traceability >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 72, \\\"traceability\\\": 30, \\\"supervision\\\": 59, \\\"contamination\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0096:noul", "group_id": "wf12:laboratory_batch:test:0096", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d8618811da.\\nC means supervision >= 41.\\nA means purity >= 71.\\nD means contamination >= 61.\\nB means traceability >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 72, \\\"traceability\\\": 30, \\\"supervision\\\": 59, \\\"contamination\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0096:uncertain", "group_id": "wf12:laboratory_batch:test:0096", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d8618811da.\\nC means supervision >= 41.\\nA means purity >= 71.\\nD means contamination >= 61.\\nB means traceability >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.8, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.13000000000000003}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0097:choice", "group_id": "wf12:laboratory_batch:test:0097", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1e2eac322e.\\nA means purity >= 89.\\nB means traceability >= 23.\\nC means supervision >= 87.\\nD means contamination >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 89, \\\"traceability\\\": 97, \\\"supervision\\\": 92, \\\"contamination\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5c38c0\": \"Outcome PROCEED.\", \"candidate_561ca0\": \"Outcome REVIEW.\", \"candidate_a479c3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5c38c0\": 0.0, \"candidate_561ca0\": 1.0, \"candidate_a479c3\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0097:score", "group_id": "wf12:laboratory_batch:test:0097", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1e2eac322e.\\nA means purity >= 89.\\nB means traceability >= 23.\\nC means supervision >= 87.\\nD means contamination >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 89, \\\"traceability\\\": 97, \\\"supervision\\\": 92, \\\"contamination\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0097:noul", "group_id": "wf12:laboratory_batch:test:0097", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1e2eac322e.\\nA means purity >= 89.\\nB means traceability >= 23.\\nC means supervision >= 87.\\nD means contamination >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 89, \\\"traceability\\\": 97, \\\"supervision\\\": 92, \\\"contamination\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0097:uncertain", "group_id": "wf12:laboratory_batch:test:0097", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1e2eac322e.\\nA means purity >= 89.\\nB means traceability >= 23.\\nC means supervision >= 87.\\nD means contamination >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.65, \\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.7725, \"true\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0098:choice", "group_id": "wf12:laboratory_batch:test:0098", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 59db34ddce.\\nA means purity >= 68.\\nD means contamination >= 87.\\nB means traceability >= 76.\\nC means supervision >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 34, \\\"traceability\\\": 86, \\\"supervision\\\": 55, \\\"contamination\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_57ec96\": \"Outcome PROCEED.\", \"candidate_289f1b\": \"Outcome REVIEW.\", \"candidate_1ead67\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_57ec96\": 0.0, \"candidate_289f1b\": 0.0, \"candidate_1ead67\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0098:score", "group_id": "wf12:laboratory_batch:test:0098", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 59db34ddce.\\nA means purity >= 68.\\nD means contamination >= 87.\\nB means traceability >= 76.\\nC means supervision >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 34, \\\"traceability\\\": 86, \\\"supervision\\\": 55, \\\"contamination\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0098:noul", "group_id": "wf12:laboratory_batch:test:0098", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 59db34ddce.\\nA means purity >= 68.\\nD means contamination >= 87.\\nB means traceability >= 76.\\nC means supervision >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 34, \\\"traceability\\\": 86, \\\"supervision\\\": 55, \\\"contamination\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0098:uncertain", "group_id": "wf12:laboratory_batch:test:0098", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 59db34ddce.\\nA means purity >= 68.\\nD means contamination >= 87.\\nB means traceability >= 76.\\nC means supervision >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 86, \\\"supervision\\\": 55, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09999999999999998, \"true\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0099:choice", "group_id": "wf12:laboratory_batch:test:0099", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 147794a33a.\\nC means supervision >= 68.\\nB means traceability >= 66.\\nA means purity >= 48.\\nD means contamination >= 82.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 87, \\\"traceability\\\": 87, \\\"supervision\\\": 86, \\\"contamination\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4aec33\": \"Outcome PROCEED.\", \"candidate_d4ab40\": \"Outcome REVIEW.\", \"candidate_d224a8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4aec33\": 1.0, \"candidate_d4ab40\": 0.0, \"candidate_d224a8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0099:score", "group_id": "wf12:laboratory_batch:test:0099", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 147794a33a.\\nC means supervision >= 68.\\nB means traceability >= 66.\\nA means purity >= 48.\\nD means contamination >= 82.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 87, \\\"traceability\\\": 87, \\\"supervision\\\": 86, \\\"contamination\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0099:noul", "group_id": "wf12:laboratory_batch:test:0099", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 147794a33a.\\nC means supervision >= 68.\\nB means traceability >= 66.\\nA means purity >= 48.\\nD means contamination >= 82.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 87, \\\"traceability\\\": 87, \\\"supervision\\\": 86, \\\"contamination\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0099:uncertain", "group_id": "wf12:laboratory_batch:test:0099", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 147794a33a.\\nC means supervision >= 68.\\nB means traceability >= 66.\\nA means purity >= 48.\\nD means contamination >= 82.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 87, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 86, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4aec33\": \"Outcome PROCEED.\", \"candidate_d4ab40\": \"Outcome REVIEW.\", \"candidate_d224a8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4aec33\": 0.8, \"candidate_d4ab40\": 0.2, \"candidate_d224a8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0100:choice", "group_id": "wf12:laboratory_batch:test:0100", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2938b3fd28.\\nA means purity >= 84.\\nC means supervision >= 42.\\nB means traceability >= 24.\\nD means contamination >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 93, \\\"traceability\\\": 9, \\\"supervision\\\": 74, \\\"contamination\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2c901c\": \"Outcome PROCEED.\", \"candidate_2b034c\": \"Outcome REVIEW.\", \"candidate_a52ce5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2c901c\": 0.0, \"candidate_2b034c\": 1.0, \"candidate_a52ce5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0100:score", "group_id": "wf12:laboratory_batch:test:0100", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2938b3fd28.\\nA means purity >= 84.\\nC means supervision >= 42.\\nB means traceability >= 24.\\nD means contamination >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 93, \\\"traceability\\\": 9, \\\"supervision\\\": 74, \\\"contamination\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0100:noul", "group_id": "wf12:laboratory_batch:test:0100", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2938b3fd28.\\nA means purity >= 84.\\nC means supervision >= 42.\\nB means traceability >= 24.\\nD means contamination >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 93, \\\"traceability\\\": 9, \\\"supervision\\\": 74, \\\"contamination\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0100:uncertain", "group_id": "wf12:laboratory_batch:test:0100", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2938b3fd28.\\nA means purity >= 84.\\nC means supervision >= 42.\\nB means traceability >= 24.\\nD means contamination >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 93, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 74, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2c901c\": \"Outcome PROCEED.\", \"candidate_2b034c\": \"Outcome REVIEW.\", \"candidate_a52ce5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2c901c\": 0.8, \"candidate_2b034c\": 0.2, \"candidate_a52ce5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0101:choice", "group_id": "wf12:laboratory_batch:test:0101", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 67ace86574.\\nB means traceability >= 89.\\nD means contamination >= 46.\\nC means supervision >= 70.\\nA means purity >= 73.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 38, \\\"traceability\\\": 90, \\\"supervision\\\": 56, \\\"contamination\\\": 66}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cf60b8\": \"Outcome PROCEED.\", \"candidate_f4a610\": \"Outcome REVIEW.\", \"candidate_c82527\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cf60b8\": 0.0, \"candidate_f4a610\": 0.0, \"candidate_c82527\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0101:score", "group_id": "wf12:laboratory_batch:test:0101", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 67ace86574.\\nB means traceability >= 89.\\nD means contamination >= 46.\\nC means supervision >= 70.\\nA means purity >= 73.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 38, \\\"traceability\\\": 90, \\\"supervision\\\": 56, \\\"contamination\\\": 66}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0101:noul", "group_id": "wf12:laboratory_batch:test:0101", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 67ace86574.\\nB means traceability >= 89.\\nD means contamination >= 46.\\nC means supervision >= 70.\\nA means purity >= 73.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 38, \\\"traceability\\\": 90, \\\"supervision\\\": 56, \\\"contamination\\\": 66}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0101:uncertain", "group_id": "wf12:laboratory_batch:test:0101", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 67ace86574.\\nB means traceability >= 89.\\nD means contamination >= 46.\\nC means supervision >= 70.\\nA means purity >= 73.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 38, \\\"traceability\\\": 90, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.1, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cf60b8\": \"Outcome PROCEED.\", \"candidate_f4a610\": \"Outcome REVIEW.\", \"candidate_c82527\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cf60b8\": 0.09000000000000001, \"candidate_f4a610\": 0.8200000000000001, \"candidate_c82527\": 0.09000000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0102:choice", "group_id": "wf12:laboratory_batch:test:0102", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5c07464108.\\nD means contamination >= 72.\\nC means supervision >= 11.\\nB means traceability >= 28.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 49, \\\"supervision\\\": 73, \\\"contamination\\\": 28}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_230922\": \"Outcome PROCEED.\", \"candidate_516242\": \"Outcome REVIEW.\", \"candidate_e021a7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_230922\": 1.0, \"candidate_516242\": 0.0, \"candidate_e021a7\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0102:score", "group_id": "wf12:laboratory_batch:test:0102", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5c07464108.\\nD means contamination >= 72.\\nC means supervision >= 11.\\nB means traceability >= 28.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 49, \\\"supervision\\\": 73, \\\"contamination\\\": 28}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0102:noul", "group_id": "wf12:laboratory_batch:test:0102", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5c07464108.\\nD means contamination >= 72.\\nC means supervision >= 11.\\nB means traceability >= 28.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 49, \\\"supervision\\\": 73, \\\"contamination\\\": 28}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0102:uncertain", "group_id": "wf12:laboratory_batch:test:0102", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5c07464108.\\nD means contamination >= 72.\\nC means supervision >= 11.\\nB means traceability >= 28.\\nA means purity >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.35, \\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.18619999999999998, \"1\": 0.3913, \"2\": 0.42250000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0103:choice", "group_id": "wf12:laboratory_batch:test:0103", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2de65cb514.\\nC means supervision >= 26.\\nB means traceability >= 72.\\nA means purity >= 86.\\nD means contamination >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 86, \\\"traceability\\\": 71, \\\"supervision\\\": 44, \\\"contamination\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a42d24\": \"Outcome PROCEED.\", \"candidate_389a28\": \"Outcome REVIEW.\", \"candidate_4dc683\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a42d24\": 0.0, \"candidate_389a28\": 1.0, \"candidate_4dc683\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0103:score", "group_id": "wf12:laboratory_batch:test:0103", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2de65cb514.\\nC means supervision >= 26.\\nB means traceability >= 72.\\nA means purity >= 86.\\nD means contamination >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 86, \\\"traceability\\\": 71, \\\"supervision\\\": 44, \\\"contamination\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0103:noul", "group_id": "wf12:laboratory_batch:test:0103", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2de65cb514.\\nC means supervision >= 26.\\nB means traceability >= 72.\\nA means purity >= 86.\\nD means contamination >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 86, \\\"traceability\\\": 71, \\\"supervision\\\": 44, \\\"contamination\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0103:uncertain", "group_id": "wf12:laboratory_batch:test:0103", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2de65cb514.\\nC means supervision >= 26.\\nB means traceability >= 72.\\nA means purity >= 86.\\nD means contamination >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 86, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 35}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.65, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0104:choice", "group_id": "wf12:laboratory_batch:test:0104", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 133fd91084.\\nA means purity >= 62.\\nC means supervision >= 13.\\nD means contamination >= 63.\\nB means traceability >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 69, \\\"traceability\\\": 52, \\\"supervision\\\": 11, \\\"contamination\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1dcc6e\": \"Outcome PROCEED.\", \"candidate_4096b9\": \"Outcome REVIEW.\", \"candidate_71b5a1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1dcc6e\": 0.0, \"candidate_4096b9\": 0.0, \"candidate_71b5a1\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0104:score", "group_id": "wf12:laboratory_batch:test:0104", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 133fd91084.\\nA means purity >= 62.\\nC means supervision >= 13.\\nD means contamination >= 63.\\nB means traceability >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 69, \\\"traceability\\\": 52, \\\"supervision\\\": 11, \\\"contamination\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0104:noul", "group_id": "wf12:laboratory_batch:test:0104", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 133fd91084.\\nA means purity >= 62.\\nC means supervision >= 13.\\nD means contamination >= 63.\\nB means traceability >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 69, \\\"traceability\\\": 52, \\\"supervision\\\": 11, \\\"contamination\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0104:uncertain", "group_id": "wf12:laboratory_batch:test:0104", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 133fd91084.\\nA means purity >= 62.\\nC means supervision >= 13.\\nD means contamination >= 63.\\nB means traceability >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 52, \\\"supervision\\\": 11, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.09999999999999998, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0105:choice", "group_id": "wf12:laboratory_batch:test:0105", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 131e9798e0.\\nB means traceability >= 58.\\nA means purity >= 12.\\nD means contamination >= 76.\\nC means supervision >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 83, \\\"supervision\\\": 81, \\\"contamination\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1446fa\": \"Outcome PROCEED.\", \"candidate_65b954\": \"Outcome REVIEW.\", \"candidate_719b5d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1446fa\": 1.0, \"candidate_65b954\": 0.0, \"candidate_719b5d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0105:score", "group_id": "wf12:laboratory_batch:test:0105", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 131e9798e0.\\nB means traceability >= 58.\\nA means purity >= 12.\\nD means contamination >= 76.\\nC means supervision >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 83, \\\"supervision\\\": 81, \\\"contamination\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0105:noul", "group_id": "wf12:laboratory_batch:test:0105", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 131e9798e0.\\nB means traceability >= 58.\\nA means purity >= 12.\\nD means contamination >= 76.\\nC means supervision >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 83, \\\"supervision\\\": 81, \\\"contamination\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0105:uncertain", "group_id": "wf12:laboratory_batch:test:0105", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 131e9798e0.\\nB means traceability >= 58.\\nA means purity >= 12.\\nD means contamination >= 76.\\nC means supervision >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 83, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.12999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0106:choice", "group_id": "wf12:laboratory_batch:test:0106", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6b7178de40.\\nA means purity >= 27.\\nD means contamination >= 48.\\nC means supervision >= 30.\\nB means traceability >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 15, \\\"traceability\\\": 73, \\\"supervision\\\": 74, \\\"contamination\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a2ca71\": \"Outcome PROCEED.\", \"candidate_16a4a8\": \"Outcome REVIEW.\", \"candidate_3c465b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a2ca71\": 0.0, \"candidate_16a4a8\": 1.0, \"candidate_3c465b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0106:score", "group_id": "wf12:laboratory_batch:test:0106", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6b7178de40.\\nA means purity >= 27.\\nD means contamination >= 48.\\nC means supervision >= 30.\\nB means traceability >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 15, \\\"traceability\\\": 73, \\\"supervision\\\": 74, \\\"contamination\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0106:noul", "group_id": "wf12:laboratory_batch:test:0106", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6b7178de40.\\nA means purity >= 27.\\nD means contamination >= 48.\\nC means supervision >= 30.\\nB means traceability >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 15, \\\"traceability\\\": 73, \\\"supervision\\\": 74, \\\"contamination\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0106:uncertain", "group_id": "wf12:laboratory_batch:test:0106", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 6b7178de40.\\nA means purity >= 27.\\nD means contamination >= 48.\\nC means supervision >= 30.\\nB means traceability >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.8, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.13}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0107:choice", "group_id": "wf12:laboratory_batch:test:0107", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 188fd2521d.\\nC means supervision >= 90.\\nD means contamination >= 75.\\nA means purity >= 87.\\nB means traceability >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 88, \\\"traceability\\\": 3, \\\"supervision\\\": 19, \\\"contamination\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ee5d96\": \"Outcome PROCEED.\", \"candidate_b55e52\": \"Outcome REVIEW.\", \"candidate_8cd052\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ee5d96\": 0.0, \"candidate_b55e52\": 0.0, \"candidate_8cd052\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0107:score", "group_id": "wf12:laboratory_batch:test:0107", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 188fd2521d.\\nC means supervision >= 90.\\nD means contamination >= 75.\\nA means purity >= 87.\\nB means traceability >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 88, \\\"traceability\\\": 3, \\\"supervision\\\": 19, \\\"contamination\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0107:noul", "group_id": "wf12:laboratory_batch:test:0107", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 188fd2521d.\\nC means supervision >= 90.\\nD means contamination >= 75.\\nA means purity >= 87.\\nB means traceability >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 88, \\\"traceability\\\": 3, \\\"supervision\\\": 19, \\\"contamination\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0107:uncertain", "group_id": "wf12:laboratory_batch:test:0107", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 188fd2521d.\\nC means supervision >= 90.\\nD means contamination >= 75.\\nA means purity >= 87.\\nB means traceability >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 3, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 95}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.95, \"true\": 0.050000000000000044}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0108:choice", "group_id": "wf12:laboratory_batch:test:0108", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 70cf5e42ee.\\nB means traceability >= 30.\\nC means supervision >= 73.\\nD means contamination >= 22.\\nA means purity >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 40, \\\"traceability\\\": 68, \\\"supervision\\\": 85, \\\"contamination\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_be9f91\": \"Outcome PROCEED.\", \"candidate_9ad787\": \"Outcome REVIEW.\", \"candidate_519061\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_be9f91\": 1.0, \"candidate_9ad787\": 0.0, \"candidate_519061\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0108:score", "group_id": "wf12:laboratory_batch:test:0108", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 70cf5e42ee.\\nB means traceability >= 30.\\nC means supervision >= 73.\\nD means contamination >= 22.\\nA means purity >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 40, \\\"traceability\\\": 68, \\\"supervision\\\": 85, \\\"contamination\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0108:noul", "group_id": "wf12:laboratory_batch:test:0108", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 70cf5e42ee.\\nB means traceability >= 30.\\nC means supervision >= 73.\\nD means contamination >= 22.\\nA means purity >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 40, \\\"traceability\\\": 68, \\\"supervision\\\": 85, \\\"contamination\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0108:uncertain", "group_id": "wf12:laboratory_batch:test:0108", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 70cf5e42ee.\\nB means traceability >= 30.\\nC means supervision >= 73.\\nD means contamination >= 22.\\nA means purity >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65, \\\"C\\\": 0.2, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_be9f91\": \"Outcome PROCEED.\", \"candidate_9ad787\": \"Outcome REVIEW.\", \"candidate_519061\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_be9f91\": 0.4304000000000001, \"candidate_9ad787\": 0.4096, \"candidate_519061\": 0.16000000000000003}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0109:choice", "group_id": "wf12:laboratory_batch:test:0109", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 362c41802a.\\nB means traceability >= 70.\\nD means contamination >= 90.\\nA means purity >= 77.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 48, \\\"traceability\\\": 61, \\\"supervision\\\": 64, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1c7c93\": \"Outcome PROCEED.\", \"candidate_e0c27c\": \"Outcome REVIEW.\", \"candidate_0a09fa\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1c7c93\": 0.0, \"candidate_e0c27c\": 1.0, \"candidate_0a09fa\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0109:score", "group_id": "wf12:laboratory_batch:test:0109", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 362c41802a.\\nB means traceability >= 70.\\nD means contamination >= 90.\\nA means purity >= 77.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 48, \\\"traceability\\\": 61, \\\"supervision\\\": 64, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0109:noul", "group_id": "wf12:laboratory_batch:test:0109", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 362c41802a.\\nB means traceability >= 70.\\nD means contamination >= 90.\\nA means purity >= 77.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 48, \\\"traceability\\\": 61, \\\"supervision\\\": 64, \\\"contamination\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0109:uncertain", "group_id": "wf12:laboratory_batch:test:0109", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 362c41802a.\\nB means traceability >= 70.\\nD means contamination >= 90.\\nA means purity >= 77.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.8, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1c7c93\": \"Outcome PROCEED.\", \"candidate_e0c27c\": \"Outcome REVIEW.\", \"candidate_0a09fa\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1c7c93\": 0.652736, \"candidate_e0c27c\": 0.21726399999999996, \"candidate_0a09fa\": 0.13}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0110:choice", "group_id": "wf12:laboratory_batch:test:0110", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8e214c8abb.\\nC means supervision >= 62.\\nA means purity >= 69.\\nB means traceability >= 63.\\nD means contamination >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 68, \\\"traceability\\\": 61, \\\"supervision\\\": 0, \\\"contamination\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dca112\": \"Outcome PROCEED.\", \"candidate_ea78c8\": \"Outcome REVIEW.\", \"candidate_819ba5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dca112\": 0.0, \"candidate_ea78c8\": 0.0, \"candidate_819ba5\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0110:score", "group_id": "wf12:laboratory_batch:test:0110", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8e214c8abb.\\nC means supervision >= 62.\\nA means purity >= 69.\\nB means traceability >= 63.\\nD means contamination >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 68, \\\"traceability\\\": 61, \\\"supervision\\\": 0, \\\"contamination\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0110:noul", "group_id": "wf12:laboratory_batch:test:0110", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8e214c8abb.\\nC means supervision >= 62.\\nA means purity >= 69.\\nB means traceability >= 63.\\nD means contamination >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 68, \\\"traceability\\\": 61, \\\"supervision\\\": 0, \\\"contamination\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0110:uncertain", "group_id": "wf12:laboratory_batch:test:0110", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8e214c8abb.\\nC means supervision >= 62.\\nA means purity >= 69.\\nB means traceability >= 63.\\nD means contamination >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 68, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 49}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.9, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dca112\": \"Outcome PROCEED.\", \"candidate_ea78c8\": \"Outcome REVIEW.\", \"candidate_819ba5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dca112\": 0.0, \"candidate_ea78c8\": 0.05, \"candidate_819ba5\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0111:choice", "group_id": "wf12:laboratory_batch:test:0111", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 15fe11f6d8.\\nA means purity >= 86.\\nB means traceability >= 81.\\nC means supervision >= 74.\\nD means contamination >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 90, \\\"traceability\\\": 85, \\\"supervision\\\": 80, \\\"contamination\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2d0d22\": \"Outcome PROCEED.\", \"candidate_511f03\": \"Outcome REVIEW.\", \"candidate_a72e9f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2d0d22\": 1.0, \"candidate_511f03\": 0.0, \"candidate_a72e9f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0111:score", "group_id": "wf12:laboratory_batch:test:0111", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 15fe11f6d8.\\nA means purity >= 86.\\nB means traceability >= 81.\\nC means supervision >= 74.\\nD means contamination >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 90, \\\"traceability\\\": 85, \\\"supervision\\\": 80, \\\"contamination\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0111:noul", "group_id": "wf12:laboratory_batch:test:0111", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 15fe11f6d8.\\nA means purity >= 86.\\nB means traceability >= 81.\\nC means supervision >= 74.\\nD means contamination >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 90, \\\"traceability\\\": 85, \\\"supervision\\\": 80, \\\"contamination\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0111:uncertain", "group_id": "wf12:laboratory_batch:test:0111", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 15fe11f6d8.\\nA means purity >= 86.\\nB means traceability >= 81.\\nC means supervision >= 74.\\nD means contamination >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.35, \\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.24250624999999998, \"1\": 0.5299937499999999, \"2\": 0.2275}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0112:choice", "group_id": "wf12:laboratory_batch:test:0112", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 620da0caf8.\\nC means supervision >= 50.\\nB means traceability >= 34.\\nA means purity >= 28.\\nD means contamination >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 96, \\\"supervision\\\": 73, \\\"contamination\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bbe22e\": \"Outcome PROCEED.\", \"candidate_8e60e2\": \"Outcome REVIEW.\", \"candidate_80c8ee\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bbe22e\": 0.0, \"candidate_8e60e2\": 1.0, \"candidate_80c8ee\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0112:score", "group_id": "wf12:laboratory_batch:test:0112", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 620da0caf8.\\nC means supervision >= 50.\\nB means traceability >= 34.\\nA means purity >= 28.\\nD means contamination >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 96, \\\"supervision\\\": 73, \\\"contamination\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0112:noul", "group_id": "wf12:laboratory_batch:test:0112", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 620da0caf8.\\nC means supervision >= 50.\\nB means traceability >= 34.\\nA means purity >= 28.\\nD means contamination >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 96, \\\"supervision\\\": 73, \\\"contamination\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0112:uncertain", "group_id": "wf12:laboratory_batch:test:0112", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 620da0caf8.\\nC means supervision >= 50.\\nB means traceability >= 34.\\nA means purity >= 28.\\nD means contamination >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 96, \\\"supervision\\\": 73, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.19999999999999996, \"1\": 0.8, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0113:choice", "group_id": "wf12:laboratory_batch:test:0113", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e680e7d232.\\nB means traceability >= 64.\\nD means contamination >= 87.\\nC means supervision >= 13.\\nA means purity >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": 93, \\\"supervision\\\": 12, \\\"contamination\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1cadb9\": \"Outcome PROCEED.\", \"candidate_5c35ec\": \"Outcome REVIEW.\", \"candidate_a6b9f1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1cadb9\": 0.0, \"candidate_5c35ec\": 0.0, \"candidate_a6b9f1\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0113:score", "group_id": "wf12:laboratory_batch:test:0113", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e680e7d232.\\nB means traceability >= 64.\\nD means contamination >= 87.\\nC means supervision >= 13.\\nA means purity >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": 93, \\\"supervision\\\": 12, \\\"contamination\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0113:noul", "group_id": "wf12:laboratory_batch:test:0113", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e680e7d232.\\nB means traceability >= 64.\\nD means contamination >= 87.\\nC means supervision >= 13.\\nA means purity >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 50, \\\"traceability\\\": 93, \\\"supervision\\\": 12, \\\"contamination\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0113:uncertain", "group_id": "wf12:laboratory_batch:test:0113", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e680e7d232.\\nB means traceability >= 64.\\nD means contamination >= 87.\\nC means supervision >= 13.\\nA means purity >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 93, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 100}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.1, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0114:choice", "group_id": "wf12:laboratory_batch:test:0114", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a8f270bf2e.\\nB means traceability >= 20.\\nD means contamination >= 34.\\nC means supervision >= 61.\\nA means purity >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 98, \\\"supervision\\\": 13, \\\"contamination\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c1145a\": \"Outcome PROCEED.\", \"candidate_ce14b3\": \"Outcome REVIEW.\", \"candidate_0be632\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c1145a\": 1.0, \"candidate_ce14b3\": 0.0, \"candidate_0be632\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0114:score", "group_id": "wf12:laboratory_batch:test:0114", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a8f270bf2e.\\nB means traceability >= 20.\\nD means contamination >= 34.\\nC means supervision >= 61.\\nA means purity >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 98, \\\"supervision\\\": 13, \\\"contamination\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0114:noul", "group_id": "wf12:laboratory_batch:test:0114", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a8f270bf2e.\\nB means traceability >= 20.\\nD means contamination >= 34.\\nC means supervision >= 61.\\nA means purity >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 98, \\\"supervision\\\": 13, \\\"contamination\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0114:uncertain", "group_id": "wf12:laboratory_batch:test:0114", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a8f270bf2e.\\nB means traceability >= 20.\\nD means contamination >= 34.\\nC means supervision >= 61.\\nA means purity >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.8, \\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8775, \"true\": 0.1225}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0115:choice", "group_id": "wf12:laboratory_batch:test:0115", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 46a7a1b4a7.\\nA means purity >= 83.\\nB means traceability >= 31.\\nD means contamination >= 78.\\nC means supervision >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 57, \\\"traceability\\\": 3, \\\"supervision\\\": 78, \\\"contamination\\\": 78}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b28e19\": \"Outcome PROCEED.\", \"candidate_bc7273\": \"Outcome REVIEW.\", \"candidate_a20ce1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b28e19\": 0.0, \"candidate_bc7273\": 1.0, \"candidate_a20ce1\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0115:score", "group_id": "wf12:laboratory_batch:test:0115", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 46a7a1b4a7.\\nA means purity >= 83.\\nB means traceability >= 31.\\nD means contamination >= 78.\\nC means supervision >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 57, \\\"traceability\\\": 3, \\\"supervision\\\": 78, \\\"contamination\\\": 78}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0115:noul", "group_id": "wf12:laboratory_batch:test:0115", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 46a7a1b4a7.\\nA means purity >= 83.\\nB means traceability >= 31.\\nD means contamination >= 78.\\nC means supervision >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 57, \\\"traceability\\\": 3, \\\"supervision\\\": 78, \\\"contamination\\\": 78}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0115:uncertain", "group_id": "wf12:laboratory_batch:test:0115", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 46a7a1b4a7.\\nA means purity >= 83.\\nB means traceability >= 31.\\nD means contamination >= 78.\\nC means supervision >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.2, \\\"C\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.72, \"true\": 0.28}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0116:choice", "group_id": "wf12:laboratory_batch:test:0116", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a7e5dc1885.\\nA means purity >= 49.\\nB means traceability >= 35.\\nD means contamination >= 81.\\nC means supervision >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 37, \\\"supervision\\\": 1, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d5a18a\": \"Outcome PROCEED.\", \"candidate_826beb\": \"Outcome REVIEW.\", \"candidate_e2ff07\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d5a18a\": 0.0, \"candidate_826beb\": 0.0, \"candidate_e2ff07\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0116:score", "group_id": "wf12:laboratory_batch:test:0116", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a7e5dc1885.\\nA means purity >= 49.\\nB means traceability >= 35.\\nD means contamination >= 81.\\nC means supervision >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 37, \\\"supervision\\\": 1, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0116:noul", "group_id": "wf12:laboratory_batch:test:0116", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a7e5dc1885.\\nA means purity >= 49.\\nB means traceability >= 35.\\nD means contamination >= 81.\\nC means supervision >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 37, \\\"supervision\\\": 1, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0116:uncertain", "group_id": "wf12:laboratory_batch:test:0116", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a7e5dc1885.\\nA means purity >= 49.\\nB means traceability >= 35.\\nD means contamination >= 81.\\nC means supervision >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 37, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.95, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9975, \"true\": 0.0025000000000000022}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0117:choice", "group_id": "wf12:laboratory_batch:test:0117", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference da0a6fd104.\\nB means traceability >= 53.\\nA means purity >= 28.\\nC means supervision >= 36.\\nD means contamination >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 87, \\\"supervision\\\": 44, \\\"contamination\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7ec70b\": \"Outcome PROCEED.\", \"candidate_8d5b57\": \"Outcome REVIEW.\", \"candidate_2b193f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7ec70b\": 1.0, \"candidate_8d5b57\": 0.0, \"candidate_2b193f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0117:score", "group_id": "wf12:laboratory_batch:test:0117", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference da0a6fd104.\\nB means traceability >= 53.\\nA means purity >= 28.\\nC means supervision >= 36.\\nD means contamination >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 87, \\\"supervision\\\": 44, \\\"contamination\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0117:noul", "group_id": "wf12:laboratory_batch:test:0117", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference da0a6fd104.\\nB means traceability >= 53.\\nA means purity >= 28.\\nC means supervision >= 36.\\nD means contamination >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 87, \\\"supervision\\\": 44, \\\"contamination\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0117:uncertain", "group_id": "wf12:laboratory_batch:test:0117", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference da0a6fd104.\\nB means traceability >= 53.\\nA means purity >= 28.\\nC means supervision >= 36.\\nD means contamination >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 87, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 72}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7ec70b\": \"Outcome PROCEED.\", \"candidate_8d5b57\": \"Outcome REVIEW.\", \"candidate_2b193f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7ec70b\": 0.48, \"candidate_8d5b57\": 0.52, \"candidate_2b193f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0118:choice", "group_id": "wf12:laboratory_batch:test:0118", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a97f12fa4b.\\nA means purity >= 75.\\nC means supervision >= 32.\\nD means contamination >= 57.\\nB means traceability >= 36.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 28, \\\"supervision\\\": 67, \\\"contamination\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_10b0fc\": \"Outcome PROCEED.\", \"candidate_895d92\": \"Outcome REVIEW.\", \"candidate_3e7470\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_10b0fc\": 0.0, \"candidate_895d92\": 1.0, \"candidate_3e7470\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0118:score", "group_id": "wf12:laboratory_batch:test:0118", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a97f12fa4b.\\nA means purity >= 75.\\nC means supervision >= 32.\\nD means contamination >= 57.\\nB means traceability >= 36.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 28, \\\"supervision\\\": 67, \\\"contamination\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0118:noul", "group_id": "wf12:laboratory_batch:test:0118", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a97f12fa4b.\\nA means purity >= 75.\\nC means supervision >= 32.\\nD means contamination >= 57.\\nB means traceability >= 36.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 28, \\\"supervision\\\": 67, \\\"contamination\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0118:uncertain", "group_id": "wf12:laboratory_batch:test:0118", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a97f12fa4b.\\nA means purity >= 75.\\nC means supervision >= 32.\\nD means contamination >= 57.\\nB means traceability >= 36.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 28, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_10b0fc\": \"Outcome PROCEED.\", \"candidate_895d92\": \"Outcome REVIEW.\", \"candidate_3e7470\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_10b0fc\": 0.187945205479452, \"candidate_895d92\": 0.5110958904109589, \"candidate_3e7470\": 0.3009589041095891}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0119:choice", "group_id": "wf12:laboratory_batch:test:0119", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3964ee4610.\\nB means traceability >= 17.\\nC means supervision >= 18.\\nD means contamination >= 89.\\nA means purity >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 59, \\\"traceability\\\": 48, \\\"supervision\\\": 1, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c1d785\": \"Outcome PROCEED.\", \"candidate_d2f348\": \"Outcome REVIEW.\", \"candidate_c68752\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c1d785\": 0.0, \"candidate_d2f348\": 0.0, \"candidate_c68752\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0119:score", "group_id": "wf12:laboratory_batch:test:0119", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3964ee4610.\\nB means traceability >= 17.\\nC means supervision >= 18.\\nD means contamination >= 89.\\nA means purity >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 59, \\\"traceability\\\": 48, \\\"supervision\\\": 1, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0119:noul", "group_id": "wf12:laboratory_batch:test:0119", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3964ee4610.\\nB means traceability >= 17.\\nC means supervision >= 18.\\nD means contamination >= 89.\\nA means purity >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 59, \\\"traceability\\\": 48, \\\"supervision\\\": 1, \\\"contamination\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0119:uncertain", "group_id": "wf12:laboratory_batch:test:0119", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3964ee4610.\\nB means traceability >= 17.\\nC means supervision >= 18.\\nD means contamination >= 89.\\nA means purity >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 48, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 89}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c1d785\": \"Outcome PROCEED.\", \"candidate_d2f348\": \"Outcome REVIEW.\", \"candidate_c68752\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c1d785\": 0.0, \"candidate_d2f348\": 0.1, \"candidate_c68752\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0120:choice", "group_id": "wf12:laboratory_batch:test:0120", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4d75f4a69b.\\nA means purity >= 55.\\nB means traceability >= 20.\\nC means supervision >= 76.\\nD means contamination >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 1, \\\"traceability\\\": 17, \\\"supervision\\\": 84, \\\"contamination\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d73153\": \"Outcome PROCEED.\", \"candidate_14827b\": \"Outcome REVIEW.\", \"candidate_2a1d38\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d73153\": 1.0, \"candidate_14827b\": 0.0, \"candidate_2a1d38\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0120:score", "group_id": "wf12:laboratory_batch:test:0120", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4d75f4a69b.\\nA means purity >= 55.\\nB means traceability >= 20.\\nC means supervision >= 76.\\nD means contamination >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 1, \\\"traceability\\\": 17, \\\"supervision\\\": 84, \\\"contamination\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0120:noul", "group_id": "wf12:laboratory_batch:test:0120", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4d75f4a69b.\\nA means purity >= 55.\\nB means traceability >= 20.\\nC means supervision >= 76.\\nD means contamination >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 1, \\\"traceability\\\": 17, \\\"supervision\\\": 84, \\\"contamination\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0120:uncertain", "group_id": "wf12:laboratory_batch:test:0120", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 4d75f4a69b.\\nA means purity >= 55.\\nB means traceability >= 20.\\nC means supervision >= 76.\\nD means contamination >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 17, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 42}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8, \"1\": 0.19999999999999996, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0121:choice", "group_id": "wf12:laboratory_batch:test:0121", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2ed1d3a45e.\\nB means traceability >= 74.\\nA means purity >= 49.\\nD means contamination >= 15.\\nC means supervision >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 25, \\\"traceability\\\": 36, \\\"supervision\\\": 99, \\\"contamination\\\": 24}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9a518b\": \"Outcome PROCEED.\", \"candidate_94c19d\": \"Outcome REVIEW.\", \"candidate_f71540\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9a518b\": 0.0, \"candidate_94c19d\": 1.0, \"candidate_f71540\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0121:score", "group_id": "wf12:laboratory_batch:test:0121", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2ed1d3a45e.\\nB means traceability >= 74.\\nA means purity >= 49.\\nD means contamination >= 15.\\nC means supervision >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 25, \\\"traceability\\\": 36, \\\"supervision\\\": 99, \\\"contamination\\\": 24}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0121:noul", "group_id": "wf12:laboratory_batch:test:0121", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2ed1d3a45e.\\nB means traceability >= 74.\\nA means purity >= 49.\\nD means contamination >= 15.\\nC means supervision >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 25, \\\"traceability\\\": 36, \\\"supervision\\\": 99, \\\"contamination\\\": 24}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0121:uncertain", "group_id": "wf12:laboratory_batch:test:0121", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2ed1d3a45e.\\nB means traceability >= 74.\\nA means purity >= 49.\\nD means contamination >= 15.\\nC means supervision >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.8, \\\"C\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.1600712328767123, \"1\": 0.5599287671232877, \"2\": 0.28}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0122:choice", "group_id": "wf12:laboratory_batch:test:0122", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 494b415a7e.\\nC means supervision >= 89.\\nD means contamination >= 26.\\nA means purity >= 61.\\nB means traceability >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 92, \\\"traceability\\\": 1, \\\"supervision\\\": 19, \\\"contamination\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d21d49\": \"Outcome PROCEED.\", \"candidate_395cce\": \"Outcome REVIEW.\", \"candidate_e04bda\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d21d49\": 0.0, \"candidate_395cce\": 0.0, \"candidate_e04bda\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0122:score", "group_id": "wf12:laboratory_batch:test:0122", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 494b415a7e.\\nC means supervision >= 89.\\nD means contamination >= 26.\\nA means purity >= 61.\\nB means traceability >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 92, \\\"traceability\\\": 1, \\\"supervision\\\": 19, \\\"contamination\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0122:noul", "group_id": "wf12:laboratory_batch:test:0122", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 494b415a7e.\\nC means supervision >= 89.\\nD means contamination >= 26.\\nA means purity >= 61.\\nB means traceability >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 92, \\\"traceability\\\": 1, \\\"supervision\\\": 19, \\\"contamination\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0122:uncertain", "group_id": "wf12:laboratory_batch:test:0122", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 494b415a7e.\\nC means supervision >= 89.\\nD means contamination >= 26.\\nA means purity >= 61.\\nB means traceability >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 92, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 19, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.095, \"1\": 0.855, \"2\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0123:choice", "group_id": "wf12:laboratory_batch:test:0123", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 06159f1d51.\\nB means traceability >= 55.\\nD means contamination >= 12.\\nC means supervision >= 23.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 67, \\\"supervision\\\": 60, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_25ec41\": \"Outcome PROCEED.\", \"candidate_20ab63\": \"Outcome REVIEW.\", \"candidate_def4f5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_25ec41\": 1.0, \"candidate_20ab63\": 0.0, \"candidate_def4f5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0123:score", "group_id": "wf12:laboratory_batch:test:0123", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 06159f1d51.\\nB means traceability >= 55.\\nD means contamination >= 12.\\nC means supervision >= 23.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 67, \\\"supervision\\\": 60, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0123:noul", "group_id": "wf12:laboratory_batch:test:0123", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 06159f1d51.\\nB means traceability >= 55.\\nD means contamination >= 12.\\nC means supervision >= 23.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 83, \\\"traceability\\\": 67, \\\"supervision\\\": 60, \\\"contamination\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0123:uncertain", "group_id": "wf12:laboratory_batch:test:0123", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 06159f1d51.\\nB means traceability >= 55.\\nD means contamination >= 12.\\nC means supervision >= 23.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.8, \\\"C\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.12999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0124:choice", "group_id": "wf12:laboratory_batch:test:0124", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c146ee8e22.\\nB means traceability >= 19.\\nC means supervision >= 13.\\nA means purity >= 23.\\nD means contamination >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 11, \\\"traceability\\\": 17, \\\"supervision\\\": 53, \\\"contamination\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c0f616\": \"Outcome PROCEED.\", \"candidate_4548c4\": \"Outcome REVIEW.\", \"candidate_01c1ba\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c0f616\": 0.0, \"candidate_4548c4\": 1.0, \"candidate_01c1ba\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0124:score", "group_id": "wf12:laboratory_batch:test:0124", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c146ee8e22.\\nB means traceability >= 19.\\nC means supervision >= 13.\\nA means purity >= 23.\\nD means contamination >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 11, \\\"traceability\\\": 17, \\\"supervision\\\": 53, \\\"contamination\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0124:noul", "group_id": "wf12:laboratory_batch:test:0124", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c146ee8e22.\\nB means traceability >= 19.\\nC means supervision >= 13.\\nA means purity >= 23.\\nD means contamination >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 11, \\\"traceability\\\": 17, \\\"supervision\\\": 53, \\\"contamination\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0124:uncertain", "group_id": "wf12:laboratory_batch:test:0124", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c146ee8e22.\\nB means traceability >= 19.\\nC means supervision >= 13.\\nA means purity >= 23.\\nD means contamination >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.65, \\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8400000000000001, \"true\": 0.15999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0125:choice", "group_id": "wf12:laboratory_batch:test:0125", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 07f70e9059.\\nB means traceability >= 53.\\nD means contamination >= 90.\\nC means supervision >= 15.\\nA means purity >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 80, \\\"supervision\\\": 5, \\\"contamination\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_251505\": \"Outcome PROCEED.\", \"candidate_8dc7f5\": \"Outcome REVIEW.\", \"candidate_5f9449\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_251505\": 0.0, \"candidate_8dc7f5\": 0.0, \"candidate_5f9449\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0125:score", "group_id": "wf12:laboratory_batch:test:0125", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 07f70e9059.\\nB means traceability >= 53.\\nD means contamination >= 90.\\nC means supervision >= 15.\\nA means purity >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 80, \\\"supervision\\\": 5, \\\"contamination\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0125:noul", "group_id": "wf12:laboratory_batch:test:0125", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 07f70e9059.\\nB means traceability >= 53.\\nD means contamination >= 90.\\nC means supervision >= 15.\\nA means purity >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 80, \\\"supervision\\\": 5, \\\"contamination\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0125:uncertain", "group_id": "wf12:laboratory_batch:test:0125", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 07f70e9059.\\nB means traceability >= 53.\\nD means contamination >= 90.\\nC means supervision >= 15.\\nA means purity >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 5, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.95, \"true\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0126:choice", "group_id": "wf12:laboratory_batch:test:0126", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9fb70c84ef.\\nA means purity >= 74.\\nD means contamination >= 85.\\nB means traceability >= 77.\\nC means supervision >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 93, \\\"traceability\\\": 79, \\\"supervision\\\": 0, \\\"contamination\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5e082a\": \"Outcome PROCEED.\", \"candidate_3627fc\": \"Outcome REVIEW.\", \"candidate_d16560\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5e082a\": 1.0, \"candidate_3627fc\": 0.0, \"candidate_d16560\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0126:score", "group_id": "wf12:laboratory_batch:test:0126", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9fb70c84ef.\\nA means purity >= 74.\\nD means contamination >= 85.\\nB means traceability >= 77.\\nC means supervision >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 93, \\\"traceability\\\": 79, \\\"supervision\\\": 0, \\\"contamination\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0126:noul", "group_id": "wf12:laboratory_batch:test:0126", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9fb70c84ef.\\nA means purity >= 74.\\nD means contamination >= 85.\\nB means traceability >= 77.\\nC means supervision >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 93, \\\"traceability\\\": 79, \\\"supervision\\\": 0, \\\"contamination\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0126:uncertain", "group_id": "wf12:laboratory_batch:test:0126", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9fb70c84ef.\\nA means purity >= 74.\\nD means contamination >= 85.\\nB means traceability >= 77.\\nC means supervision >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 93, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 0, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5e082a\": \"Outcome PROCEED.\", \"candidate_3627fc\": \"Outcome REVIEW.\", \"candidate_d16560\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5e082a\": 0.13, \"candidate_3627fc\": 0.52, \"candidate_d16560\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0127:choice", "group_id": "wf12:laboratory_batch:test:0127", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1de81ca2eb.\\nA means purity >= 71.\\nD means contamination >= 66.\\nB means traceability >= 89.\\nC means supervision >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 75, \\\"traceability\\\": 93, \\\"supervision\\\": 78, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eee3a7\": \"Outcome PROCEED.\", \"candidate_139db8\": \"Outcome REVIEW.\", \"candidate_399904\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eee3a7\": 0.0, \"candidate_139db8\": 1.0, \"candidate_399904\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0127:score", "group_id": "wf12:laboratory_batch:test:0127", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1de81ca2eb.\\nA means purity >= 71.\\nD means contamination >= 66.\\nB means traceability >= 89.\\nC means supervision >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 75, \\\"traceability\\\": 93, \\\"supervision\\\": 78, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0127:noul", "group_id": "wf12:laboratory_batch:test:0127", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1de81ca2eb.\\nA means purity >= 71.\\nD means contamination >= 66.\\nB means traceability >= 89.\\nC means supervision >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 75, \\\"traceability\\\": 93, \\\"supervision\\\": 78, \\\"contamination\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0127:uncertain", "group_id": "wf12:laboratory_batch:test:0127", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1de81ca2eb.\\nA means purity >= 71.\\nD means contamination >= 66.\\nB means traceability >= 89.\\nC means supervision >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 75, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 67}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eee3a7\": \"Outcome PROCEED.\", \"candidate_139db8\": \"Outcome REVIEW.\", \"candidate_399904\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eee3a7\": 0.0, \"candidate_139db8\": 0.65, \"candidate_399904\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0128:choice", "group_id": "wf12:laboratory_batch:test:0128", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3d7b244d8f.\\nD means contamination >= 64.\\nC means supervision >= 87.\\nA means purity >= 25.\\nB means traceability >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 73, \\\"traceability\\\": 79, \\\"supervision\\\": 7, \\\"contamination\\\": 77}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_81af22\": \"Outcome PROCEED.\", \"candidate_db3156\": \"Outcome REVIEW.\", \"candidate_f7c6e9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_81af22\": 0.0, \"candidate_db3156\": 0.0, \"candidate_f7c6e9\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0128:score", "group_id": "wf12:laboratory_batch:test:0128", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3d7b244d8f.\\nD means contamination >= 64.\\nC means supervision >= 87.\\nA means purity >= 25.\\nB means traceability >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 73, \\\"traceability\\\": 79, \\\"supervision\\\": 7, \\\"contamination\\\": 77}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0128:noul", "group_id": "wf12:laboratory_batch:test:0128", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3d7b244d8f.\\nD means contamination >= 64.\\nC means supervision >= 87.\\nA means purity >= 25.\\nB means traceability >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 73, \\\"traceability\\\": 79, \\\"supervision\\\": 7, \\\"contamination\\\": 77}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0128:uncertain", "group_id": "wf12:laboratory_batch:test:0128", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 3d7b244d8f.\\nD means contamination >= 64.\\nC means supervision >= 87.\\nA means purity >= 25.\\nB means traceability >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 73, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 77}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"C\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_81af22\": \"Outcome PROCEED.\", \"candidate_db3156\": \"Outcome REVIEW.\", \"candidate_f7c6e9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_81af22\": 0.0, \"candidate_db3156\": 0.9, \"candidate_f7c6e9\": 0.09999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0129:choice", "group_id": "wf12:laboratory_batch:test:0129", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e93b51263.\\nD means contamination >= 17.\\nC means supervision >= 10.\\nA means purity >= 65.\\nB means traceability >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 38, \\\"traceability\\\": 77, \\\"supervision\\\": 48, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a8d5d7\": \"Outcome PROCEED.\", \"candidate_78a42f\": \"Outcome REVIEW.\", \"candidate_568656\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a8d5d7\": 1.0, \"candidate_78a42f\": 0.0, \"candidate_568656\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0129:score", "group_id": "wf12:laboratory_batch:test:0129", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e93b51263.\\nD means contamination >= 17.\\nC means supervision >= 10.\\nA means purity >= 65.\\nB means traceability >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 38, \\\"traceability\\\": 77, \\\"supervision\\\": 48, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0129:noul", "group_id": "wf12:laboratory_batch:test:0129", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e93b51263.\\nD means contamination >= 17.\\nC means supervision >= 10.\\nA means purity >= 65.\\nB means traceability >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 38, \\\"traceability\\\": 77, \\\"supervision\\\": 48, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0129:uncertain", "group_id": "wf12:laboratory_batch:test:0129", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e93b51263.\\nD means contamination >= 17.\\nC means supervision >= 10.\\nA means purity >= 65.\\nB means traceability >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 38, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 48, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.19999999999999996, \"1\": 0.8, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0130:choice", "group_id": "wf12:laboratory_batch:test:0130", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference afed64d56a.\\nB means traceability >= 67.\\nD means contamination >= 51.\\nC means supervision >= 84.\\nA means purity >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 47, \\\"supervision\\\": 90, \\\"contamination\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_18c41e\": \"Outcome PROCEED.\", \"candidate_4df209\": \"Outcome REVIEW.\", \"candidate_4f97ad\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_18c41e\": 0.0, \"candidate_4df209\": 1.0, \"candidate_4f97ad\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0130:score", "group_id": "wf12:laboratory_batch:test:0130", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference afed64d56a.\\nB means traceability >= 67.\\nD means contamination >= 51.\\nC means supervision >= 84.\\nA means purity >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 47, \\\"supervision\\\": 90, \\\"contamination\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0130:noul", "group_id": "wf12:laboratory_batch:test:0130", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference afed64d56a.\\nB means traceability >= 67.\\nD means contamination >= 51.\\nC means supervision >= 84.\\nA means purity >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 47, \\\"supervision\\\": 90, \\\"contamination\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0130:uncertain", "group_id": "wf12:laboratory_batch:test:0130", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference afed64d56a.\\nB means traceability >= 67.\\nD means contamination >= 51.\\nC means supervision >= 84.\\nA means purity >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 18, \\\"traceability\\\": 47, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.7168000000000001, \"1\": 0.2623999999999999, \"2\": 0.020799999999999975}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0131:choice", "group_id": "wf12:laboratory_batch:test:0131", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference cfb1b217fe.\\nB means traceability >= 41.\\nD means contamination >= 26.\\nC means supervision >= 68.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": 98, \\\"supervision\\\": 14, \\\"contamination\\\": 40}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c8be26\": \"Outcome PROCEED.\", \"candidate_75dd5a\": \"Outcome REVIEW.\", \"candidate_b82858\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c8be26\": 0.0, \"candidate_75dd5a\": 0.0, \"candidate_b82858\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0131:score", "group_id": "wf12:laboratory_batch:test:0131", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference cfb1b217fe.\\nB means traceability >= 41.\\nD means contamination >= 26.\\nC means supervision >= 68.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": 98, \\\"supervision\\\": 14, \\\"contamination\\\": 40}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0131:noul", "group_id": "wf12:laboratory_batch:test:0131", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference cfb1b217fe.\\nB means traceability >= 41.\\nD means contamination >= 26.\\nC means supervision >= 68.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": 98, \\\"supervision\\\": 14, \\\"contamination\\\": 40}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0131:uncertain", "group_id": "wf12:laboratory_batch:test:0131", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference cfb1b217fe.\\nB means traceability >= 41.\\nD means contamination >= 26.\\nC means supervision >= 68.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 98, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 40}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.1, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0132:choice", "group_id": "wf12:laboratory_batch:test:0132", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e129d748f5.\\nC means supervision >= 43.\\nD means contamination >= 36.\\nA means purity >= 47.\\nB means traceability >= 20.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 98, \\\"traceability\\\": 40, \\\"supervision\\\": 18, \\\"contamination\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a774f\": \"Outcome PROCEED.\", \"candidate_d84bab\": \"Outcome REVIEW.\", \"candidate_34c7a1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a774f\": 1.0, \"candidate_d84bab\": 0.0, \"candidate_34c7a1\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0132:score", "group_id": "wf12:laboratory_batch:test:0132", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e129d748f5.\\nC means supervision >= 43.\\nD means contamination >= 36.\\nA means purity >= 47.\\nB means traceability >= 20.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 98, \\\"traceability\\\": 40, \\\"supervision\\\": 18, \\\"contamination\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0132:noul", "group_id": "wf12:laboratory_batch:test:0132", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e129d748f5.\\nC means supervision >= 43.\\nD means contamination >= 36.\\nA means purity >= 47.\\nB means traceability >= 20.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 98, \\\"traceability\\\": 40, \\\"supervision\\\": 18, \\\"contamination\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0132:uncertain", "group_id": "wf12:laboratory_batch:test:0132", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e129d748f5.\\nC means supervision >= 43.\\nD means contamination >= 36.\\nA means purity >= 47.\\nB means traceability >= 20.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.35, \\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.7725, \"true\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0133:choice", "group_id": "wf12:laboratory_batch:test:0133", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ea089d88db.\\nA means purity >= 15.\\nB means traceability >= 42.\\nC means supervision >= 81.\\nD means contamination >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 20, \\\"supervision\\\": 89, \\\"contamination\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_97c2d6\": \"Outcome PROCEED.\", \"candidate_6e7488\": \"Outcome REVIEW.\", \"candidate_ac72e2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_97c2d6\": 0.0, \"candidate_6e7488\": 1.0, \"candidate_ac72e2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0133:score", "group_id": "wf12:laboratory_batch:test:0133", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ea089d88db.\\nA means purity >= 15.\\nB means traceability >= 42.\\nC means supervision >= 81.\\nD means contamination >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 20, \\\"supervision\\\": 89, \\\"contamination\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0133:noul", "group_id": "wf12:laboratory_batch:test:0133", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ea089d88db.\\nA means purity >= 15.\\nB means traceability >= 42.\\nC means supervision >= 81.\\nD means contamination >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 7, \\\"traceability\\\": 20, \\\"supervision\\\": 89, \\\"contamination\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0133:uncertain", "group_id": "wf12:laboratory_batch:test:0133", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ea089d88db.\\nA means purity >= 15.\\nB means traceability >= 42.\\nC means supervision >= 81.\\nD means contamination >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 20, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 100}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0134:choice", "group_id": "wf12:laboratory_batch:test:0134", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2e7967b05b.\\nC means supervision >= 85.\\nA means purity >= 18.\\nB means traceability >= 38.\\nD means contamination >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 74, \\\"traceability\\\": 14, \\\"supervision\\\": 47, \\\"contamination\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_70fdef\": \"Outcome PROCEED.\", \"candidate_2092ce\": \"Outcome REVIEW.\", \"candidate_9b11ae\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_70fdef\": 0.0, \"candidate_2092ce\": 0.0, \"candidate_9b11ae\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0134:score", "group_id": "wf12:laboratory_batch:test:0134", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2e7967b05b.\\nC means supervision >= 85.\\nA means purity >= 18.\\nB means traceability >= 38.\\nD means contamination >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 74, \\\"traceability\\\": 14, \\\"supervision\\\": 47, \\\"contamination\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0134:noul", "group_id": "wf12:laboratory_batch:test:0134", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2e7967b05b.\\nC means supervision >= 85.\\nA means purity >= 18.\\nB means traceability >= 38.\\nD means contamination >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 74, \\\"traceability\\\": 14, \\\"supervision\\\": 47, \\\"contamination\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0134:uncertain", "group_id": "wf12:laboratory_batch:test:0134", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2e7967b05b.\\nC means supervision >= 85.\\nA means purity >= 18.\\nB means traceability >= 38.\\nD means contamination >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 74, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 47, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09999999999999998, \"true\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0135:choice", "group_id": "wf12:laboratory_batch:test:0135", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8718d15089.\\nC means supervision >= 88.\\nA means purity >= 38.\\nD means contamination >= 54.\\nB means traceability >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 6, \\\"supervision\\\": 94, \\\"contamination\\\": 48}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bde3bb\": \"Outcome PROCEED.\", \"candidate_dbfd33\": \"Outcome REVIEW.\", \"candidate_2ea31b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bde3bb\": 1.0, \"candidate_dbfd33\": 0.0, \"candidate_2ea31b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0135:score", "group_id": "wf12:laboratory_batch:test:0135", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8718d15089.\\nC means supervision >= 88.\\nA means purity >= 38.\\nD means contamination >= 54.\\nB means traceability >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 6, \\\"supervision\\\": 94, \\\"contamination\\\": 48}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0135:noul", "group_id": "wf12:laboratory_batch:test:0135", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8718d15089.\\nC means supervision >= 88.\\nA means purity >= 38.\\nD means contamination >= 54.\\nB means traceability >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 6, \\\"supervision\\\": 94, \\\"contamination\\\": 48}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0135:uncertain", "group_id": "wf12:laboratory_batch:test:0135", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 8718d15089.\\nC means supervision >= 88.\\nA means purity >= 38.\\nD means contamination >= 54.\\nB means traceability >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 4, \\\"traceability\\\": 6, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.2, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bde3bb\": \"Outcome PROCEED.\", \"candidate_dbfd33\": \"Outcome REVIEW.\", \"candidate_2ea31b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bde3bb\": 0.06999999999999999, \"candidate_dbfd33\": 0.41, \"candidate_2ea31b\": 0.52}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0136:choice", "group_id": "wf12:laboratory_batch:test:0136", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 434b0b704f.\\nB means traceability >= 45.\\nC means supervision >= 61.\\nD means contamination >= 81.\\nA means purity >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": 23, \\\"supervision\\\": 83, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3e52e7\": \"Outcome PROCEED.\", \"candidate_2ae15c\": \"Outcome REVIEW.\", \"candidate_e5a8d1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3e52e7\": 0.0, \"candidate_2ae15c\": 1.0, \"candidate_e5a8d1\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0136:score", "group_id": "wf12:laboratory_batch:test:0136", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 434b0b704f.\\nB means traceability >= 45.\\nC means supervision >= 61.\\nD means contamination >= 81.\\nA means purity >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": 23, \\\"supervision\\\": 83, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0136:noul", "group_id": "wf12:laboratory_batch:test:0136", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 434b0b704f.\\nB means traceability >= 45.\\nC means supervision >= 61.\\nD means contamination >= 81.\\nA means purity >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": 23, \\\"supervision\\\": 83, \\\"contamination\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0136:uncertain", "group_id": "wf12:laboratory_batch:test:0136", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 434b0b704f.\\nB means traceability >= 45.\\nC means supervision >= 61.\\nD means contamination >= 81.\\nA means purity >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 83, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3e52e7\": \"Outcome PROCEED.\", \"candidate_2ae15c\": \"Outcome REVIEW.\", \"candidate_e5a8d1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3e52e7\": 0.65, \"candidate_2ae15c\": 0.35, \"candidate_e5a8d1\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0137:choice", "group_id": "wf12:laboratory_batch:test:0137", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 830ebb79d5.\\nC means supervision >= 50.\\nB means traceability >= 82.\\nA means purity >= 59.\\nD means contamination >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 19, \\\"traceability\\\": 15, \\\"supervision\\\": 11, \\\"contamination\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1b5da5\": \"Outcome PROCEED.\", \"candidate_93e422\": \"Outcome REVIEW.\", \"candidate_1fd8e7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1b5da5\": 0.0, \"candidate_93e422\": 0.0, \"candidate_1fd8e7\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0137:score", "group_id": "wf12:laboratory_batch:test:0137", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 830ebb79d5.\\nC means supervision >= 50.\\nB means traceability >= 82.\\nA means purity >= 59.\\nD means contamination >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 19, \\\"traceability\\\": 15, \\\"supervision\\\": 11, \\\"contamination\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0137:noul", "group_id": "wf12:laboratory_batch:test:0137", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 830ebb79d5.\\nC means supervision >= 50.\\nB means traceability >= 82.\\nA means purity >= 59.\\nD means contamination >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 19, \\\"traceability\\\": 15, \\\"supervision\\\": 11, \\\"contamination\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0137:uncertain", "group_id": "wf12:laboratory_batch:test:0137", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 830ebb79d5.\\nC means supervision >= 50.\\nB means traceability >= 82.\\nA means purity >= 59.\\nD means contamination >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 19, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 76}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1b5da5\": \"Outcome PROCEED.\", \"candidate_93e422\": \"Outcome REVIEW.\", \"candidate_1fd8e7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1b5da5\": 0.0, \"candidate_93e422\": 0.1, \"candidate_1fd8e7\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0138:choice", "group_id": "wf12:laboratory_batch:test:0138", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c655e3e5cf.\\nD means contamination >= 44.\\nB means traceability >= 37.\\nC means supervision >= 28.\\nA means purity >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 17, \\\"traceability\\\": 52, \\\"supervision\\\": 45, \\\"contamination\\\": 40}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4d3bbf\": \"Outcome PROCEED.\", \"candidate_d9e668\": \"Outcome REVIEW.\", \"candidate_2ab9cd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4d3bbf\": 1.0, \"candidate_d9e668\": 0.0, \"candidate_2ab9cd\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0138:score", "group_id": "wf12:laboratory_batch:test:0138", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c655e3e5cf.\\nD means contamination >= 44.\\nB means traceability >= 37.\\nC means supervision >= 28.\\nA means purity >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 17, \\\"traceability\\\": 52, \\\"supervision\\\": 45, \\\"contamination\\\": 40}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0138:noul", "group_id": "wf12:laboratory_batch:test:0138", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c655e3e5cf.\\nD means contamination >= 44.\\nB means traceability >= 37.\\nC means supervision >= 28.\\nA means purity >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 17, \\\"traceability\\\": 52, \\\"supervision\\\": 45, \\\"contamination\\\": 40}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0138:uncertain", "group_id": "wf12:laboratory_batch:test:0138", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference c655e3e5cf.\\nD means contamination >= 44.\\nB means traceability >= 37.\\nC means supervision >= 28.\\nA means purity >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 52, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 40}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9299999999999999, \"1\": 0.06999999999999998, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0139:choice", "group_id": "wf12:laboratory_batch:test:0139", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 55489b92e3.\\nC means supervision >= 74.\\nB means traceability >= 48.\\nA means purity >= 74.\\nD means contamination >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 82, \\\"traceability\\\": 41, \\\"supervision\\\": 26, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_98d57e\": \"Outcome PROCEED.\", \"candidate_1ceae8\": \"Outcome REVIEW.\", \"candidate_87984a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_98d57e\": 0.0, \"candidate_1ceae8\": 1.0, \"candidate_87984a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0139:score", "group_id": "wf12:laboratory_batch:test:0139", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 55489b92e3.\\nC means supervision >= 74.\\nB means traceability >= 48.\\nA means purity >= 74.\\nD means contamination >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 82, \\\"traceability\\\": 41, \\\"supervision\\\": 26, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0139:noul", "group_id": "wf12:laboratory_batch:test:0139", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 55489b92e3.\\nC means supervision >= 74.\\nB means traceability >= 48.\\nA means purity >= 74.\\nD means contamination >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 82, \\\"traceability\\\": 41, \\\"supervision\\\": 26, \\\"contamination\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0139:uncertain", "group_id": "wf12:laboratory_batch:test:0139", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 55489b92e3.\\nC means supervision >= 74.\\nB means traceability >= 48.\\nA means purity >= 74.\\nD means contamination >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 82, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 26, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.34904109589041094, \"1\": 0.3009589041095891, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0140:choice", "group_id": "wf12:laboratory_batch:test:0140", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ee0ec8c4eb.\\nA means purity >= 30.\\nB means traceability >= 26.\\nC means supervision >= 90.\\nD means contamination >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 75, \\\"traceability\\\": 24, \\\"supervision\\\": 88, \\\"contamination\\\": 52}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1381cb\": \"Outcome PROCEED.\", \"candidate_148eb1\": \"Outcome REVIEW.\", \"candidate_63d365\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1381cb\": 0.0, \"candidate_148eb1\": 0.0, \"candidate_63d365\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0140:score", "group_id": "wf12:laboratory_batch:test:0140", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ee0ec8c4eb.\\nA means purity >= 30.\\nB means traceability >= 26.\\nC means supervision >= 90.\\nD means contamination >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 75, \\\"traceability\\\": 24, \\\"supervision\\\": 88, \\\"contamination\\\": 52}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0140:noul", "group_id": "wf12:laboratory_batch:test:0140", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ee0ec8c4eb.\\nA means purity >= 30.\\nB means traceability >= 26.\\nC means supervision >= 90.\\nD means contamination >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 75, \\\"traceability\\\": 24, \\\"supervision\\\": 88, \\\"contamination\\\": 52}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0140:uncertain", "group_id": "wf12:laboratory_batch:test:0140", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference ee0ec8c4eb.\\nA means purity >= 30.\\nB means traceability >= 26.\\nC means supervision >= 90.\\nD means contamination >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.05, \\\"C\\\": 0.1, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.005112500000000003, \"1\": 0.13988750000000005, \"2\": 0.8550000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0141:choice", "group_id": "wf12:laboratory_batch:test:0141", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e0ddc9dd5.\\nC means supervision >= 84.\\nD means contamination >= 42.\\nB means traceability >= 55.\\nA means purity >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 1, \\\"traceability\\\": 37, \\\"supervision\\\": 100, \\\"contamination\\\": 38}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3bfad9\": \"Outcome PROCEED.\", \"candidate_5e434a\": \"Outcome REVIEW.\", \"candidate_91617c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3bfad9\": 1.0, \"candidate_5e434a\": 0.0, \"candidate_91617c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0141:score", "group_id": "wf12:laboratory_batch:test:0141", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e0ddc9dd5.\\nC means supervision >= 84.\\nD means contamination >= 42.\\nB means traceability >= 55.\\nA means purity >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 1, \\\"traceability\\\": 37, \\\"supervision\\\": 100, \\\"contamination\\\": 38}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0141:noul", "group_id": "wf12:laboratory_batch:test:0141", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e0ddc9dd5.\\nC means supervision >= 84.\\nD means contamination >= 42.\\nB means traceability >= 55.\\nA means purity >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 1, \\\"traceability\\\": 37, \\\"supervision\\\": 100, \\\"contamination\\\": 38}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0141:uncertain", "group_id": "wf12:laboratory_batch:test:0141", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 5e0ddc9dd5.\\nC means supervision >= 84.\\nD means contamination >= 42.\\nB means traceability >= 55.\\nA means purity >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.35, \\\"C\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.12999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0142:choice", "group_id": "wf12:laboratory_batch:test:0142", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1b3ab6960b.\\nD means contamination >= 17.\\nA means purity >= 34.\\nB means traceability >= 49.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 46, \\\"traceability\\\": 22, \\\"supervision\\\": 94, \\\"contamination\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_645034\": \"Outcome PROCEED.\", \"candidate_66e366\": \"Outcome REVIEW.\", \"candidate_e08618\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_645034\": 0.0, \"candidate_66e366\": 1.0, \"candidate_e08618\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0142:score", "group_id": "wf12:laboratory_batch:test:0142", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1b3ab6960b.\\nD means contamination >= 17.\\nA means purity >= 34.\\nB means traceability >= 49.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 46, \\\"traceability\\\": 22, \\\"supervision\\\": 94, \\\"contamination\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0142:noul", "group_id": "wf12:laboratory_batch:test:0142", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1b3ab6960b.\\nD means contamination >= 17.\\nA means purity >= 34.\\nB means traceability >= 49.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 46, \\\"traceability\\\": 22, \\\"supervision\\\": 94, \\\"contamination\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0142:uncertain", "group_id": "wf12:laboratory_batch:test:0142", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 1b3ab6960b.\\nD means contamination >= 17.\\nA means purity >= 34.\\nB means traceability >= 49.\\nC means supervision >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.65, \\\"C\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.3599999999999999, \"true\": 0.6400000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0143:choice", "group_id": "wf12:laboratory_batch:test:0143", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 56ec45fb60.\\nB means traceability >= 61.\\nD means contamination >= 46.\\nC means supervision >= 39.\\nA means purity >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 42, \\\"traceability\\\": 60, \\\"supervision\\\": 38, \\\"contamination\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e9bd85\": \"Outcome PROCEED.\", \"candidate_597c1a\": \"Outcome REVIEW.\", \"candidate_e2bfa9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e9bd85\": 0.0, \"candidate_597c1a\": 0.0, \"candidate_e2bfa9\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0143:score", "group_id": "wf12:laboratory_batch:test:0143", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 56ec45fb60.\\nB means traceability >= 61.\\nD means contamination >= 46.\\nC means supervision >= 39.\\nA means purity >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 42, \\\"traceability\\\": 60, \\\"supervision\\\": 38, \\\"contamination\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0143:noul", "group_id": "wf12:laboratory_batch:test:0143", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 56ec45fb60.\\nB means traceability >= 61.\\nD means contamination >= 46.\\nC means supervision >= 39.\\nA means purity >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 42, \\\"traceability\\\": 60, \\\"supervision\\\": 38, \\\"contamination\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0143:uncertain", "group_id": "wf12:laboratory_batch:test:0143", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 56ec45fb60.\\nB means traceability >= 61.\\nD means contamination >= 46.\\nC means supervision >= 39.\\nA means purity >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 42, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 85}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.050000000000000044, \"true\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0144:choice", "group_id": "wf12:laboratory_batch:test:0144", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 85d9f8484c.\\nC means supervision >= 12.\\nD means contamination >= 25.\\nB means traceability >= 43.\\nA means purity >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 68, \\\"supervision\\\": 6, \\\"contamination\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e6a28e\": \"Outcome PROCEED.\", \"candidate_24b121\": \"Outcome REVIEW.\", \"candidate_9bc7e9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e6a28e\": 1.0, \"candidate_24b121\": 0.0, \"candidate_9bc7e9\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0144:score", "group_id": "wf12:laboratory_batch:test:0144", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 85d9f8484c.\\nC means supervision >= 12.\\nD means contamination >= 25.\\nB means traceability >= 43.\\nA means purity >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 68, \\\"supervision\\\": 6, \\\"contamination\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0144:noul", "group_id": "wf12:laboratory_batch:test:0144", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 85d9f8484c.\\nC means supervision >= 12.\\nD means contamination >= 25.\\nB means traceability >= 43.\\nA means purity >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 68, \\\"supervision\\\": 6, \\\"contamination\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0144:uncertain", "group_id": "wf12:laboratory_batch:test:0144", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 85d9f8484c.\\nC means supervision >= 12.\\nD means contamination >= 25.\\nB means traceability >= 43.\\nA means purity >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 68, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e6a28e\": \"Outcome PROCEED.\", \"candidate_24b121\": \"Outcome REVIEW.\", \"candidate_9bc7e9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e6a28e\": 0.8, \"candidate_24b121\": 0.13, \"candidate_9bc7e9\": 0.06999999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0145:choice", "group_id": "wf12:laboratory_batch:test:0145", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a8ac93ac6a.\\nD means contamination >= 30.\\nC means supervision >= 46.\\nB means traceability >= 72.\\nA means purity >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 98, \\\"supervision\\\": 61, \\\"contamination\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_127c6d\": \"Outcome PROCEED.\", \"candidate_40df74\": \"Outcome REVIEW.\", \"candidate_996ce5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_127c6d\": 0.0, \"candidate_40df74\": 1.0, \"candidate_996ce5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0145:score", "group_id": "wf12:laboratory_batch:test:0145", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a8ac93ac6a.\\nD means contamination >= 30.\\nC means supervision >= 46.\\nB means traceability >= 72.\\nA means purity >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 98, \\\"supervision\\\": 61, \\\"contamination\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0145:noul", "group_id": "wf12:laboratory_batch:test:0145", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a8ac93ac6a.\\nD means contamination >= 30.\\nC means supervision >= 46.\\nB means traceability >= 72.\\nA means purity >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 98, \\\"supervision\\\": 61, \\\"contamination\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0145:uncertain", "group_id": "wf12:laboratory_batch:test:0145", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference a8ac93ac6a.\\nD means contamination >= 30.\\nC means supervision >= 46.\\nB means traceability >= 72.\\nA means purity >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 10, \\\"traceability\\\": 98, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_127c6d\": \"Outcome PROCEED.\", \"candidate_40df74\": \"Outcome REVIEW.\", \"candidate_996ce5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_127c6d\": 0.7168000000000001, \"candidate_40df74\": 0.2623999999999999, \"candidate_996ce5\": 0.020799999999999975}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0146:choice", "group_id": "wf12:laboratory_batch:test:0146", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 52d96a9141.\\nD means contamination >= 21.\\nC means supervision >= 44.\\nB means traceability >= 34.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 32, \\\"traceability\\\": 8, \\\"supervision\\\": 34, \\\"contamination\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c888e6\": \"Outcome PROCEED.\", \"candidate_a512f2\": \"Outcome REVIEW.\", \"candidate_680d9d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c888e6\": 0.0, \"candidate_a512f2\": 0.0, \"candidate_680d9d\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0146:score", "group_id": "wf12:laboratory_batch:test:0146", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 52d96a9141.\\nD means contamination >= 21.\\nC means supervision >= 44.\\nB means traceability >= 34.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 32, \\\"traceability\\\": 8, \\\"supervision\\\": 34, \\\"contamination\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0146:noul", "group_id": "wf12:laboratory_batch:test:0146", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 52d96a9141.\\nD means contamination >= 21.\\nC means supervision >= 44.\\nB means traceability >= 34.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 32, \\\"traceability\\\": 8, \\\"supervision\\\": 34, \\\"contamination\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0146:uncertain", "group_id": "wf12:laboratory_batch:test:0146", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 52d96a9141.\\nD means contamination >= 21.\\nC means supervision >= 44.\\nB means traceability >= 34.\\nA means purity >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 32, \\\"traceability\\\": 8, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.05, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c888e6\": \"Outcome PROCEED.\", \"candidate_a512f2\": \"Outcome REVIEW.\", \"candidate_680d9d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c888e6\": 0.004999999999999999, \"candidate_a512f2\": 0.13999999999999999, \"candidate_680d9d\": 0.855}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0147:choice", "group_id": "wf12:laboratory_batch:test:0147", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 822563d401.\\nA means purity >= 17.\\nD means contamination >= 56.\\nC means supervision >= 38.\\nB means traceability >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 91, \\\"supervision\\\": 92, \\\"contamination\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3ceaeb\": \"Outcome PROCEED.\", \"candidate_6ca666\": \"Outcome REVIEW.\", \"candidate_20cb94\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3ceaeb\": 1.0, \"candidate_6ca666\": 0.0, \"candidate_20cb94\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0147:score", "group_id": "wf12:laboratory_batch:test:0147", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 822563d401.\\nA means purity >= 17.\\nD means contamination >= 56.\\nC means supervision >= 38.\\nB means traceability >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 91, \\\"supervision\\\": 92, \\\"contamination\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0147:noul", "group_id": "wf12:laboratory_batch:test:0147", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 822563d401.\\nA means purity >= 17.\\nD means contamination >= 56.\\nC means supervision >= 38.\\nB means traceability >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 85, \\\"traceability\\\": 91, \\\"supervision\\\": 92, \\\"contamination\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0147:uncertain", "group_id": "wf12:laboratory_batch:test:0147", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 822563d401.\\nA means purity >= 17.\\nD means contamination >= 56.\\nC means supervision >= 38.\\nB means traceability >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 91, \\\"supervision\\\": 92, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.65, \"1\": 0.35, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0148:choice", "group_id": "wf12:laboratory_batch:test:0148", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9318996976.\\nA means purity >= 35.\\nB means traceability >= 43.\\nD means contamination >= 87.\\nC means supervision >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": 11, \\\"supervision\\\": 42, \\\"contamination\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7622e1\": \"Outcome PROCEED.\", \"candidate_807b92\": \"Outcome REVIEW.\", \"candidate_e2cac3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7622e1\": 0.0, \"candidate_807b92\": 1.0, \"candidate_e2cac3\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0148:score", "group_id": "wf12:laboratory_batch:test:0148", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9318996976.\\nA means purity >= 35.\\nB means traceability >= 43.\\nD means contamination >= 87.\\nC means supervision >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": 11, \\\"supervision\\\": 42, \\\"contamination\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0148:noul", "group_id": "wf12:laboratory_batch:test:0148", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9318996976.\\nA means purity >= 35.\\nB means traceability >= 43.\\nD means contamination >= 87.\\nC means supervision >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": 11, \\\"supervision\\\": 42, \\\"contamination\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0148:uncertain", "group_id": "wf12:laboratory_batch:test:0148", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 9318996976.\\nA means purity >= 35.\\nB means traceability >= 43.\\nD means contamination >= 87.\\nC means supervision >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 30, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 42, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.65, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0149:choice", "group_id": "wf12:laboratory_batch:test:0149", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e416add14e.\\nC means supervision >= 81.\\nA means purity >= 63.\\nD means contamination >= 38.\\nB means traceability >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 52, \\\"supervision\\\": 46, \\\"contamination\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_07cdbb\": \"Outcome PROCEED.\", \"candidate_e7ee5c\": \"Outcome REVIEW.\", \"candidate_dbbb1f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_07cdbb\": 0.0, \"candidate_e7ee5c\": 0.0, \"candidate_dbbb1f\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0149:score", "group_id": "wf12:laboratory_batch:test:0149", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e416add14e.\\nC means supervision >= 81.\\nA means purity >= 63.\\nD means contamination >= 38.\\nB means traceability >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 52, \\\"supervision\\\": 46, \\\"contamination\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0149:noul", "group_id": "wf12:laboratory_batch:test:0149", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e416add14e.\\nC means supervision >= 81.\\nA means purity >= 63.\\nD means contamination >= 38.\\nB means traceability >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 52, \\\"supervision\\\": 46, \\\"contamination\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0149:uncertain", "group_id": "wf12:laboratory_batch:test:0149", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference e416add14e.\\nC means supervision >= 81.\\nA means purity >= 63.\\nD means contamination >= 38.\\nB means traceability >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 99, \\\"traceability\\\": 52, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.05, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.95, \"1\": 0.0025000000000000005, \"2\": 0.0475}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0150:choice", "group_id": "wf12:laboratory_batch:test:0150", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 29b9b067d8.\\nD means contamination >= 10.\\nA means purity >= 73.\\nB means traceability >= 75.\\nC means supervision >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 77, \\\"traceability\\\": 13, \\\"supervision\\\": 92, \\\"contamination\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_438d1e\": \"Outcome PROCEED.\", \"candidate_09708f\": \"Outcome REVIEW.\", \"candidate_6d7387\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_438d1e\": 1.0, \"candidate_09708f\": 0.0, \"candidate_6d7387\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0150:score", "group_id": "wf12:laboratory_batch:test:0150", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 29b9b067d8.\\nD means contamination >= 10.\\nA means purity >= 73.\\nB means traceability >= 75.\\nC means supervision >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 77, \\\"traceability\\\": 13, \\\"supervision\\\": 92, \\\"contamination\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0150:noul", "group_id": "wf12:laboratory_batch:test:0150", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 29b9b067d8.\\nD means contamination >= 10.\\nA means purity >= 73.\\nB means traceability >= 75.\\nC means supervision >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 77, \\\"traceability\\\": 13, \\\"supervision\\\": 92, \\\"contamination\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0150:uncertain", "group_id": "wf12:laboratory_batch:test:0150", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 29b9b067d8.\\nD means contamination >= 10.\\nA means purity >= 73.\\nB means traceability >= 75.\\nC means supervision >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.8, \\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8775000000000001, \"true\": 0.12249999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0151:choice", "group_id": "wf12:laboratory_batch:test:0151", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2581fae7f6.\\nC means supervision >= 28.\\nD means contamination >= 82.\\nA means purity >= 34.\\nB means traceability >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 15, \\\"traceability\\\": 3, \\\"supervision\\\": 5, \\\"contamination\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bcc826\": \"Outcome PROCEED.\", \"candidate_3af0f6\": \"Outcome REVIEW.\", \"candidate_d88b31\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bcc826\": 0.0, \"candidate_3af0f6\": 1.0, \"candidate_d88b31\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0151:score", "group_id": "wf12:laboratory_batch:test:0151", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2581fae7f6.\\nC means supervision >= 28.\\nD means contamination >= 82.\\nA means purity >= 34.\\nB means traceability >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 15, \\\"traceability\\\": 3, \\\"supervision\\\": 5, \\\"contamination\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0151:noul", "group_id": "wf12:laboratory_batch:test:0151", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2581fae7f6.\\nC means supervision >= 28.\\nD means contamination >= 82.\\nA means purity >= 34.\\nB means traceability >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 15, \\\"traceability\\\": 3, \\\"supervision\\\": 5, \\\"contamination\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0151:uncertain", "group_id": "wf12:laboratory_batch:test:0151", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 2581fae7f6.\\nC means supervision >= 28.\\nD means contamination >= 82.\\nA means purity >= 34.\\nB means traceability >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 3, \\\"supervision\\\": 5, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0152:choice", "group_id": "wf12:laboratory_batch:test:0152", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b60d924945.\\nC means supervision >= 56.\\nA means purity >= 70.\\nD means contamination >= 39.\\nB means traceability >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 6, \\\"traceability\\\": 10, \\\"supervision\\\": 44, \\\"contamination\\\": 64}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7cd314\": \"Outcome PROCEED.\", \"candidate_115071\": \"Outcome REVIEW.\", \"candidate_fdde4f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7cd314\": 0.0, \"candidate_115071\": 0.0, \"candidate_fdde4f\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0152:score", "group_id": "wf12:laboratory_batch:test:0152", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b60d924945.\\nC means supervision >= 56.\\nA means purity >= 70.\\nD means contamination >= 39.\\nB means traceability >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 6, \\\"traceability\\\": 10, \\\"supervision\\\": 44, \\\"contamination\\\": 64}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0152:noul", "group_id": "wf12:laboratory_batch:test:0152", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b60d924945.\\nC means supervision >= 56.\\nA means purity >= 70.\\nD means contamination >= 39.\\nB means traceability >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 6, \\\"traceability\\\": 10, \\\"supervision\\\": 44, \\\"contamination\\\": 64}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0152:uncertain", "group_id": "wf12:laboratory_batch:test:0152", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference b60d924945.\\nC means supervision >= 56.\\nA means purity >= 70.\\nD means contamination >= 39.\\nB means traceability >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 10, \\\"supervision\\\": 44, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9, \"true\": 0.1}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0153:choice", "group_id": "wf12:laboratory_batch:test:0153", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 278f8ce5f1.\\nC means supervision >= 45.\\nD means contamination >= 24.\\nB means traceability >= 81.\\nA means purity >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 82, \\\"supervision\\\": 58, \\\"contamination\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0baf70\": \"Outcome PROCEED.\", \"candidate_b8710a\": \"Outcome REVIEW.\", \"candidate_636f31\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0baf70\": 1.0, \"candidate_b8710a\": 0.0, \"candidate_636f31\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0153:score", "group_id": "wf12:laboratory_batch:test:0153", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 278f8ce5f1.\\nC means supervision >= 45.\\nD means contamination >= 24.\\nB means traceability >= 81.\\nA means purity >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 82, \\\"supervision\\\": 58, \\\"contamination\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0153:noul", "group_id": "wf12:laboratory_batch:test:0153", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 278f8ce5f1.\\nC means supervision >= 45.\\nD means contamination >= 24.\\nB means traceability >= 81.\\nA means purity >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 91, \\\"traceability\\\": 82, \\\"supervision\\\": 58, \\\"contamination\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0153:uncertain", "group_id": "wf12:laboratory_batch:test:0153", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 278f8ce5f1.\\nC means supervision >= 45.\\nD means contamination >= 24.\\nB means traceability >= 81.\\nA means purity >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 82, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": 4}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"C\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0baf70\": \"Outcome PROCEED.\", \"candidate_b8710a\": \"Outcome REVIEW.\", \"candidate_636f31\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0baf70\": 0.8400000000000002, \"candidate_b8710a\": 0.15999999999999998, \"candidate_636f31\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0154:choice", "group_id": "wf12:laboratory_batch:test:0154", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 297cd2eefd.\\nD means contamination >= 63.\\nA means purity >= 47.\\nB means traceability >= 61.\\nC means supervision >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 16, \\\"traceability\\\": 80, \\\"supervision\\\": 4, \\\"contamination\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e3b5a9\": \"Outcome PROCEED.\", \"candidate_1a8ca9\": \"Outcome REVIEW.\", \"candidate_459eca\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e3b5a9\": 0.0, \"candidate_1a8ca9\": 1.0, \"candidate_459eca\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0154:score", "group_id": "wf12:laboratory_batch:test:0154", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 297cd2eefd.\\nD means contamination >= 63.\\nA means purity >= 47.\\nB means traceability >= 61.\\nC means supervision >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 16, \\\"traceability\\\": 80, \\\"supervision\\\": 4, \\\"contamination\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0154:noul", "group_id": "wf12:laboratory_batch:test:0154", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 297cd2eefd.\\nD means contamination >= 63.\\nA means purity >= 47.\\nB means traceability >= 61.\\nC means supervision >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 16, \\\"traceability\\\": 80, \\\"supervision\\\": 4, \\\"contamination\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0154:uncertain", "group_id": "wf12:laboratory_batch:test:0154", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 297cd2eefd.\\nD means contamination >= 63.\\nA means purity >= 47.\\nB means traceability >= 61.\\nC means supervision >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 16, \\\"traceability\\\": 80, \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e3b5a9\": \"Outcome PROCEED.\", \"candidate_1a8ca9\": \"Outcome REVIEW.\", \"candidate_459eca\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e3b5a9\": 0.31360000000000005, \"candidate_1a8ca9\": 0.6188000000000001, \"candidate_459eca\": 0.06759999999999992}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0155:choice", "group_id": "wf12:laboratory_batch:test:0155", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 03fd186aea.\\nD means contamination >= 34.\\nA means purity >= 50.\\nC means supervision >= 56.\\nB means traceability >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 27, \\\"traceability\\\": 15, \\\"supervision\\\": 35, \\\"contamination\\\": 63}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1c0e08\": \"Outcome PROCEED.\", \"candidate_1b55a9\": \"Outcome REVIEW.\", \"candidate_5f80cf\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1c0e08\": 0.0, \"candidate_1b55a9\": 0.0, \"candidate_5f80cf\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0155:score", "group_id": "wf12:laboratory_batch:test:0155", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 03fd186aea.\\nD means contamination >= 34.\\nA means purity >= 50.\\nC means supervision >= 56.\\nB means traceability >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 27, \\\"traceability\\\": 15, \\\"supervision\\\": 35, \\\"contamination\\\": 63}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0155:noul", "group_id": "wf12:laboratory_batch:test:0155", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 03fd186aea.\\nD means contamination >= 34.\\nA means purity >= 50.\\nC means supervision >= 56.\\nB means traceability >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 27, \\\"traceability\\\": 15, \\\"supervision\\\": 35, \\\"contamination\\\": 63}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0155:uncertain", "group_id": "wf12:laboratory_batch:test:0155", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 03fd186aea.\\nD means contamination >= 34.\\nA means purity >= 50.\\nC means supervision >= 56.\\nB means traceability >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.1, \\\"C\\\": 0.95, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1c0e08\": \"Outcome PROCEED.\", \"candidate_1b55a9\": \"Outcome REVIEW.\", \"candidate_5f80cf\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1c0e08\": 0.09504999999999998, \"candidate_1b55a9\": 0.8599499999999998, \"candidate_5f80cf\": 0.04500000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0156:choice", "group_id": "wf12:laboratory_batch:test:0156", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 79a96f6c76.\\nA means purity >= 14.\\nD means contamination >= 50.\\nB means traceability >= 59.\\nC means supervision >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 29, \\\"traceability\\\": 19, \\\"supervision\\\": 42, \\\"contamination\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2f00b2\": \"Outcome PROCEED.\", \"candidate_1455e0\": \"Outcome REVIEW.\", \"candidate_04376d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2f00b2\": 1.0, \"candidate_1455e0\": 0.0, \"candidate_04376d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0156:score", "group_id": "wf12:laboratory_batch:test:0156", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 79a96f6c76.\\nA means purity >= 14.\\nD means contamination >= 50.\\nB means traceability >= 59.\\nC means supervision >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 29, \\\"traceability\\\": 19, \\\"supervision\\\": 42, \\\"contamination\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0156:noul", "group_id": "wf12:laboratory_batch:test:0156", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 79a96f6c76.\\nA means purity >= 14.\\nD means contamination >= 50.\\nB means traceability >= 59.\\nC means supervision >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 29, \\\"traceability\\\": 19, \\\"supervision\\\": 42, \\\"contamination\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0156:uncertain", "group_id": "wf12:laboratory_batch:test:0156", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 79a96f6c76.\\nA means purity >= 14.\\nD means contamination >= 50.\\nB means traceability >= 59.\\nC means supervision >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 29, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 42, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.19999999999999996, \"1\": 0.8, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0157:choice", "group_id": "wf12:laboratory_batch:test:0157", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 05bdfbd0f6.\\nB means traceability >= 87.\\nA means purity >= 17.\\nD means contamination >= 30.\\nC means supervision >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 67, \\\"traceability\\\": 83, \\\"supervision\\\": 36, \\\"contamination\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4e7884\": \"Outcome PROCEED.\", \"candidate_087378\": \"Outcome REVIEW.\", \"candidate_769771\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4e7884\": 0.0, \"candidate_087378\": 1.0, \"candidate_769771\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0157:score", "group_id": "wf12:laboratory_batch:test:0157", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 05bdfbd0f6.\\nB means traceability >= 87.\\nA means purity >= 17.\\nD means contamination >= 30.\\nC means supervision >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 67, \\\"traceability\\\": 83, \\\"supervision\\\": 36, \\\"contamination\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0157:noul", "group_id": "wf12:laboratory_batch:test:0157", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 05bdfbd0f6.\\nB means traceability >= 87.\\nA means purity >= 17.\\nD means contamination >= 30.\\nC means supervision >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 67, \\\"traceability\\\": 83, \\\"supervision\\\": 36, \\\"contamination\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0157:uncertain", "group_id": "wf12:laboratory_batch:test:0157", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 05bdfbd0f6.\\nB means traceability >= 87.\\nA means purity >= 17.\\nD means contamination >= 30.\\nC means supervision >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 67, \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": 36, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.10739726027397256, \"1\": 0.0926027397260274, \"2\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0158:choice", "group_id": "wf12:laboratory_batch:test:0158", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d8affe2743.\\nC means supervision >= 10.\\nD means contamination >= 16.\\nA means purity >= 26.\\nB means traceability >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 1, \\\"traceability\\\": 55, \\\"supervision\\\": 0, \\\"contamination\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ff2e32\": \"Outcome PROCEED.\", \"candidate_92d5d6\": \"Outcome REVIEW.\", \"candidate_edfa6e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ff2e32\": 0.0, \"candidate_92d5d6\": 0.0, \"candidate_edfa6e\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0158:score", "group_id": "wf12:laboratory_batch:test:0158", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d8affe2743.\\nC means supervision >= 10.\\nD means contamination >= 16.\\nA means purity >= 26.\\nB means traceability >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 1, \\\"traceability\\\": 55, \\\"supervision\\\": 0, \\\"contamination\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0158:noul", "group_id": "wf12:laboratory_batch:test:0158", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d8affe2743.\\nC means supervision >= 10.\\nD means contamination >= 16.\\nA means purity >= 26.\\nB means traceability >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 1, \\\"traceability\\\": 55, \\\"supervision\\\": 0, \\\"contamination\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0158:uncertain", "group_id": "wf12:laboratory_batch:test:0158", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference d8affe2743.\\nC means supervision >= 10.\\nD means contamination >= 16.\\nA means purity >= 26.\\nB means traceability >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": 55, \\\"supervision\\\": 0, \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.09999999999999998, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0159:choice", "group_id": "wf12:laboratory_batch:test:0159", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0d95e1c575.\\nC means supervision >= 55.\\nD means contamination >= 28.\\nA means purity >= 75.\\nB means traceability >= 82.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 94, \\\"supervision\\\": 7, \\\"contamination\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5ada65\": \"Outcome PROCEED.\", \"candidate_2fdcb7\": \"Outcome REVIEW.\", \"candidate_bb87a5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5ada65\": 1.0, \"candidate_2fdcb7\": 0.0, \"candidate_bb87a5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0159:score", "group_id": "wf12:laboratory_batch:test:0159", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0d95e1c575.\\nC means supervision >= 55.\\nD means contamination >= 28.\\nA means purity >= 75.\\nB means traceability >= 82.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 94, \\\"supervision\\\": 7, \\\"contamination\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0159:noul", "group_id": "wf12:laboratory_batch:test:0159", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0d95e1c575.\\nC means supervision >= 55.\\nD means contamination >= 28.\\nA means purity >= 75.\\nB means traceability >= 82.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": 78, \\\"traceability\\\": 94, \\\"supervision\\\": 7, \\\"contamination\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:laboratory_batch:test:0159:uncertain", "group_id": "wf12:laboratory_batch:test:0159", "split": "test", "family": "laboratory_batch", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: laboratory batch. Case reference 0d95e1c575.\\nC means supervision >= 55.\\nD means contamination >= 28.\\nA means purity >= 75.\\nB means traceability >= 82.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND NOT C) THEN HOLD.\\n2. IF ((A OR C) AND (B OR C) AND NOT D) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"purity\\\": \\\"unobserved\\\", \\\"traceability\\\": \\\"unobserved\\\", \\\"supervision\\\": \\\"unobserved\\\", \\\"contamination\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.65, \\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.7725, \"true\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0000:choice", "group_id": "wf12:archive_loan:test:0000", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cffdfe555d.\\nB means insurance >= 72.\\nA means condition >= 73.\\nD means restriction >= 15.\\nC means authorization >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 94, \\\"authorization\\\": 100, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bb45a2\": \"Outcome PROCEED.\", \"candidate_64d20a\": \"Outcome REVIEW.\", \"candidate_f0ed03\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bb45a2\": 1.0, \"candidate_64d20a\": 0.0, \"candidate_f0ed03\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0000:score", "group_id": "wf12:archive_loan:test:0000", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cffdfe555d.\\nB means insurance >= 72.\\nA means condition >= 73.\\nD means restriction >= 15.\\nC means authorization >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 94, \\\"authorization\\\": 100, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0000:noul", "group_id": "wf12:archive_loan:test:0000", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cffdfe555d.\\nB means insurance >= 72.\\nA means condition >= 73.\\nD means restriction >= 15.\\nC means authorization >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 94, \\\"authorization\\\": 100, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0000:uncertain", "group_id": "wf12:archive_loan:test:0000", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cffdfe555d.\\nB means insurance >= 72.\\nA means condition >= 73.\\nD means restriction >= 15.\\nC means authorization >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 100, \\\"restriction\\\": 32}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.65, \"1\": 0.0, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0001:choice", "group_id": "wf12:archive_loan:test:0001", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference db00f2ad8d.\\nC means authorization >= 34.\\nA means condition >= 14.\\nB means insurance >= 89.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 25, \\\"authorization\\\": 64, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_70f1c4\": \"Outcome PROCEED.\", \"candidate_4b2945\": \"Outcome REVIEW.\", \"candidate_6c81f2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_70f1c4\": 0.0, \"candidate_4b2945\": 1.0, \"candidate_6c81f2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0001:score", "group_id": "wf12:archive_loan:test:0001", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference db00f2ad8d.\\nC means authorization >= 34.\\nA means condition >= 14.\\nB means insurance >= 89.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 25, \\\"authorization\\\": 64, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0001:noul", "group_id": "wf12:archive_loan:test:0001", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference db00f2ad8d.\\nC means authorization >= 34.\\nA means condition >= 14.\\nB means insurance >= 89.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 25, \\\"authorization\\\": 64, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0001:uncertain", "group_id": "wf12:archive_loan:test:0001", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference db00f2ad8d.\\nC means authorization >= 34.\\nA means condition >= 14.\\nB means insurance >= 89.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 64, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.1, \"1\": 0.81, \"2\": 0.09000000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0002:choice", "group_id": "wf12:archive_loan:test:0002", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2527b9d7b6.\\nD means restriction >= 90.\\nB means insurance >= 56.\\nC means authorization >= 71.\\nA means condition >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 2, \\\"insurance\\\": 45, \\\"authorization\\\": 85, \\\"restriction\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7b0a8e\": \"Outcome PROCEED.\", \"candidate_453aab\": \"Outcome REVIEW.\", \"candidate_718f2f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7b0a8e\": 0.0, \"candidate_453aab\": 0.0, \"candidate_718f2f\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0002:score", "group_id": "wf12:archive_loan:test:0002", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2527b9d7b6.\\nD means restriction >= 90.\\nB means insurance >= 56.\\nC means authorization >= 71.\\nA means condition >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 2, \\\"insurance\\\": 45, \\\"authorization\\\": 85, \\\"restriction\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0002:noul", "group_id": "wf12:archive_loan:test:0002", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2527b9d7b6.\\nD means restriction >= 90.\\nB means insurance >= 56.\\nC means authorization >= 71.\\nA means condition >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 2, \\\"insurance\\\": 45, \\\"authorization\\\": 85, \\\"restriction\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0002:uncertain", "group_id": "wf12:archive_loan:test:0002", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2527b9d7b6.\\nD means restriction >= 90.\\nB means insurance >= 56.\\nC means authorization >= 71.\\nA means condition >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 45, \\\"authorization\\\": 85, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.06999999999999999, \"1\": 0.27999999999999997, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0003:choice", "group_id": "wf12:archive_loan:test:0003", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5ad9d93c39.\\nB means insurance >= 22.\\nC means authorization >= 14.\\nA means condition >= 39.\\nD means restriction >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 63, \\\"insurance\\\": 0, \\\"authorization\\\": 91, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3b3030\": \"Outcome PROCEED.\", \"candidate_0823fe\": \"Outcome REVIEW.\", \"candidate_9e847b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3b3030\": 1.0, \"candidate_0823fe\": 0.0, \"candidate_9e847b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0003:score", "group_id": "wf12:archive_loan:test:0003", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5ad9d93c39.\\nB means insurance >= 22.\\nC means authorization >= 14.\\nA means condition >= 39.\\nD means restriction >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 63, \\\"insurance\\\": 0, \\\"authorization\\\": 91, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0003:noul", "group_id": "wf12:archive_loan:test:0003", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5ad9d93c39.\\nB means insurance >= 22.\\nC means authorization >= 14.\\nA means condition >= 39.\\nD means restriction >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 63, \\\"insurance\\\": 0, \\\"authorization\\\": 91, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0003:uncertain", "group_id": "wf12:archive_loan:test:0003", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5ad9d93c39.\\nB means insurance >= 22.\\nC means authorization >= 14.\\nA means condition >= 39.\\nD means restriction >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 0, \\\"authorization\\\": 91, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8, \"true\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0004:choice", "group_id": "wf12:archive_loan:test:0004", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 215373cde2.\\nC means authorization >= 39.\\nA means condition >= 57.\\nD means restriction >= 26.\\nB means insurance >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 69, \\\"insurance\\\": 51, \\\"authorization\\\": 20, \\\"restriction\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_829ec5\": \"Outcome PROCEED.\", \"candidate_306f72\": \"Outcome REVIEW.\", \"candidate_e1d923\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_829ec5\": 0.0, \"candidate_306f72\": 1.0, \"candidate_e1d923\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0004:score", "group_id": "wf12:archive_loan:test:0004", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 215373cde2.\\nC means authorization >= 39.\\nA means condition >= 57.\\nD means restriction >= 26.\\nB means insurance >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 69, \\\"insurance\\\": 51, \\\"authorization\\\": 20, \\\"restriction\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0004:noul", "group_id": "wf12:archive_loan:test:0004", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 215373cde2.\\nC means authorization >= 39.\\nA means condition >= 57.\\nD means restriction >= 26.\\nB means insurance >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 69, \\\"insurance\\\": 51, \\\"authorization\\\": 20, \\\"restriction\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0004:uncertain", "group_id": "wf12:archive_loan:test:0004", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 215373cde2.\\nC means authorization >= 39.\\nA means condition >= 57.\\nD means restriction >= 26.\\nB means insurance >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.1, \\\"C\\\": 0.9, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.909, \"true\": 0.09100000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0005:choice", "group_id": "wf12:archive_loan:test:0005", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ada2dfa02a.\\nA means condition >= 39.\\nB means insurance >= 40.\\nD means restriction >= 10.\\nC means authorization >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": 6, \\\"authorization\\\": 71, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1914b2\": \"Outcome PROCEED.\", \"candidate_81abbd\": \"Outcome REVIEW.\", \"candidate_fa46b5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1914b2\": 0.0, \"candidate_81abbd\": 0.0, \"candidate_fa46b5\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0005:score", "group_id": "wf12:archive_loan:test:0005", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ada2dfa02a.\\nA means condition >= 39.\\nB means insurance >= 40.\\nD means restriction >= 10.\\nC means authorization >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": 6, \\\"authorization\\\": 71, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0005:noul", "group_id": "wf12:archive_loan:test:0005", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ada2dfa02a.\\nA means condition >= 39.\\nB means insurance >= 40.\\nD means restriction >= 10.\\nC means authorization >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": 6, \\\"authorization\\\": 71, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0005:uncertain", "group_id": "wf12:archive_loan:test:0005", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ada2dfa02a.\\nA means condition >= 39.\\nB means insurance >= 40.\\nD means restriction >= 10.\\nC means authorization >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 71, \\\"restriction\\\": 14}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.19999999999999984, \"true\": 0.8000000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0006:choice", "group_id": "wf12:archive_loan:test:0006", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2e63a6a13d.\\nD means restriction >= 79.\\nB means insurance >= 41.\\nC means authorization >= 36.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 65, \\\"insurance\\\": 13, \\\"authorization\\\": 64, \\\"restriction\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_137e23\": \"Outcome PROCEED.\", \"candidate_c4eb42\": \"Outcome REVIEW.\", \"candidate_9719f8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_137e23\": 1.0, \"candidate_c4eb42\": 0.0, \"candidate_9719f8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0006:score", "group_id": "wf12:archive_loan:test:0006", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2e63a6a13d.\\nD means restriction >= 79.\\nB means insurance >= 41.\\nC means authorization >= 36.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 65, \\\"insurance\\\": 13, \\\"authorization\\\": 64, \\\"restriction\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0006:noul", "group_id": "wf12:archive_loan:test:0006", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2e63a6a13d.\\nD means restriction >= 79.\\nB means insurance >= 41.\\nC means authorization >= 36.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 65, \\\"insurance\\\": 13, \\\"authorization\\\": 64, \\\"restriction\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0006:uncertain", "group_id": "wf12:archive_loan:test:0006", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2e63a6a13d.\\nD means restriction >= 79.\\nB means insurance >= 41.\\nC means authorization >= 36.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 64, \\\"restriction\\\": 23}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_137e23\": \"Outcome PROCEED.\", \"candidate_c4eb42\": \"Outcome REVIEW.\", \"candidate_9719f8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_137e23\": 0.8699999999999999, \"candidate_c4eb42\": 0.12999999999999998, \"candidate_9719f8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0007:choice", "group_id": "wf12:archive_loan:test:0007", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7ab6383a61.\\nB means insurance >= 40.\\nC means authorization >= 86.\\nD means restriction >= 30.\\nA means condition >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 75, \\\"authorization\\\": 10, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ac4986\": \"Outcome PROCEED.\", \"candidate_31e042\": \"Outcome REVIEW.\", \"candidate_dafebf\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ac4986\": 0.0, \"candidate_31e042\": 1.0, \"candidate_dafebf\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0007:score", "group_id": "wf12:archive_loan:test:0007", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7ab6383a61.\\nB means insurance >= 40.\\nC means authorization >= 86.\\nD means restriction >= 30.\\nA means condition >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 75, \\\"authorization\\\": 10, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0007:noul", "group_id": "wf12:archive_loan:test:0007", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7ab6383a61.\\nB means insurance >= 40.\\nC means authorization >= 86.\\nD means restriction >= 30.\\nA means condition >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 75, \\\"authorization\\\": 10, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0007:uncertain", "group_id": "wf12:archive_loan:test:0007", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7ab6383a61.\\nB means insurance >= 40.\\nC means authorization >= 86.\\nD means restriction >= 30.\\nA means condition >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 75, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.9, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ac4986\": \"Outcome PROCEED.\", \"candidate_31e042\": \"Outcome REVIEW.\", \"candidate_dafebf\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ac4986\": 0.9, \"candidate_31e042\": 0.09499999999999997, \"candidate_dafebf\": 0.004999999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0008:choice", "group_id": "wf12:archive_loan:test:0008", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9df0ae759e.\\nD means restriction >= 10.\\nB means insurance >= 75.\\nC means authorization >= 78.\\nA means condition >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 99, \\\"insurance\\\": 25, \\\"authorization\\\": 64, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d530db\": \"Outcome PROCEED.\", \"candidate_775eb6\": \"Outcome REVIEW.\", \"candidate_fc8486\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d530db\": 0.0, \"candidate_775eb6\": 0.0, \"candidate_fc8486\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0008:score", "group_id": "wf12:archive_loan:test:0008", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9df0ae759e.\\nD means restriction >= 10.\\nB means insurance >= 75.\\nC means authorization >= 78.\\nA means condition >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 99, \\\"insurance\\\": 25, \\\"authorization\\\": 64, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0008:noul", "group_id": "wf12:archive_loan:test:0008", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9df0ae759e.\\nD means restriction >= 10.\\nB means insurance >= 75.\\nC means authorization >= 78.\\nA means condition >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 99, \\\"insurance\\\": 25, \\\"authorization\\\": 64, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0008:uncertain", "group_id": "wf12:archive_loan:test:0008", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9df0ae759e.\\nD means restriction >= 10.\\nB means insurance >= 75.\\nC means authorization >= 78.\\nA means condition >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 99, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 64, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d530db\": \"Outcome PROCEED.\", \"candidate_775eb6\": \"Outcome REVIEW.\", \"candidate_fc8486\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d530db\": 0.0, \"candidate_775eb6\": 0.35, \"candidate_fc8486\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0009:choice", "group_id": "wf12:archive_loan:test:0009", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 66feec5a6d.\\nB means insurance >= 46.\\nD means restriction >= 33.\\nC means authorization >= 43.\\nA means condition >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 80, \\\"authorization\\\": 92, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_04985f\": \"Outcome PROCEED.\", \"candidate_ed94ea\": \"Outcome REVIEW.\", \"candidate_6a5775\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_04985f\": 1.0, \"candidate_ed94ea\": 0.0, \"candidate_6a5775\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0009:score", "group_id": "wf12:archive_loan:test:0009", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 66feec5a6d.\\nB means insurance >= 46.\\nD means restriction >= 33.\\nC means authorization >= 43.\\nA means condition >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 80, \\\"authorization\\\": 92, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0009:noul", "group_id": "wf12:archive_loan:test:0009", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 66feec5a6d.\\nB means insurance >= 46.\\nD means restriction >= 33.\\nC means authorization >= 43.\\nA means condition >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 80, \\\"authorization\\\": 92, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0009:uncertain", "group_id": "wf12:archive_loan:test:0009", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 66feec5a6d.\\nB means insurance >= 46.\\nD means restriction >= 33.\\nC means authorization >= 43.\\nA means condition >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 80, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8960000000000001, \"1\": 0.03639999999999995, \"2\": 0.06759999999999992}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0010:choice", "group_id": "wf12:archive_loan:test:0010", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fca700c0d0.\\nB means insurance >= 28.\\nC means authorization >= 32.\\nA means condition >= 39.\\nD means restriction >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 85, \\\"insurance\\\": 66, \\\"authorization\\\": 31, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_019a3d\": \"Outcome PROCEED.\", \"candidate_e91ed3\": \"Outcome REVIEW.\", \"candidate_c2486e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_019a3d\": 0.0, \"candidate_e91ed3\": 1.0, \"candidate_c2486e\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0010:score", "group_id": "wf12:archive_loan:test:0010", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fca700c0d0.\\nB means insurance >= 28.\\nC means authorization >= 32.\\nA means condition >= 39.\\nD means restriction >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 85, \\\"insurance\\\": 66, \\\"authorization\\\": 31, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0010:noul", "group_id": "wf12:archive_loan:test:0010", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fca700c0d0.\\nB means insurance >= 28.\\nC means authorization >= 32.\\nA means condition >= 39.\\nD means restriction >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 85, \\\"insurance\\\": 66, \\\"authorization\\\": 31, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0010:uncertain", "group_id": "wf12:archive_loan:test:0010", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fca700c0d0.\\nB means insurance >= 28.\\nC means authorization >= 32.\\nA means condition >= 39.\\nD means restriction >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 85, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 31, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.9, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.050000000000000044, \"2\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0011:choice", "group_id": "wf12:archive_loan:test:0011", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6570d834dd.\\nB means insurance >= 59.\\nD means restriction >= 22.\\nC means authorization >= 49.\\nA means condition >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 53, \\\"insurance\\\": 65, \\\"authorization\\\": 44, \\\"restriction\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_754896\": \"Outcome PROCEED.\", \"candidate_55416a\": \"Outcome REVIEW.\", \"candidate_b4d91c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_754896\": 0.0, \"candidate_55416a\": 0.0, \"candidate_b4d91c\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0011:score", "group_id": "wf12:archive_loan:test:0011", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6570d834dd.\\nB means insurance >= 59.\\nD means restriction >= 22.\\nC means authorization >= 49.\\nA means condition >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 53, \\\"insurance\\\": 65, \\\"authorization\\\": 44, \\\"restriction\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0011:noul", "group_id": "wf12:archive_loan:test:0011", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6570d834dd.\\nB means insurance >= 59.\\nD means restriction >= 22.\\nC means authorization >= 49.\\nA means condition >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 53, \\\"insurance\\\": 65, \\\"authorization\\\": 44, \\\"restriction\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0011:uncertain", "group_id": "wf12:archive_loan:test:0011", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6570d834dd.\\nB means insurance >= 59.\\nD means restriction >= 22.\\nC means authorization >= 49.\\nA means condition >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 53, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 44, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.35, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0012:choice", "group_id": "wf12:archive_loan:test:0012", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0f4d1a3864.\\nB means insurance >= 41.\\nC means authorization >= 79.\\nD means restriction >= 66.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 62, \\\"authorization\\\": 99, \\\"restriction\\\": 78}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a236ae\": \"Outcome PROCEED.\", \"candidate_0bc4c7\": \"Outcome REVIEW.\", \"candidate_1a67b9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a236ae\": 1.0, \"candidate_0bc4c7\": 0.0, \"candidate_1a67b9\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0012:score", "group_id": "wf12:archive_loan:test:0012", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0f4d1a3864.\\nB means insurance >= 41.\\nC means authorization >= 79.\\nD means restriction >= 66.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 62, \\\"authorization\\\": 99, \\\"restriction\\\": 78}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0012:noul", "group_id": "wf12:archive_loan:test:0012", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0f4d1a3864.\\nB means insurance >= 41.\\nC means authorization >= 79.\\nD means restriction >= 66.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 62, \\\"authorization\\\": 99, \\\"restriction\\\": 78}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0012:uncertain", "group_id": "wf12:archive_loan:test:0012", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0f4d1a3864.\\nB means insurance >= 41.\\nC means authorization >= 79.\\nD means restriction >= 66.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 62, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.7725, \"true\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0013:choice", "group_id": "wf12:archive_loan:test:0013", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f234a1f6f9.\\nD means restriction >= 52.\\nC means authorization >= 39.\\nA means condition >= 73.\\nB means insurance >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 29, \\\"authorization\\\": 37, \\\"restriction\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_213150\": \"Outcome PROCEED.\", \"candidate_5e2f86\": \"Outcome REVIEW.\", \"candidate_aeea5f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_213150\": 0.0, \"candidate_5e2f86\": 1.0, \"candidate_aeea5f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0013:score", "group_id": "wf12:archive_loan:test:0013", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f234a1f6f9.\\nD means restriction >= 52.\\nC means authorization >= 39.\\nA means condition >= 73.\\nB means insurance >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 29, \\\"authorization\\\": 37, \\\"restriction\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0013:noul", "group_id": "wf12:archive_loan:test:0013", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f234a1f6f9.\\nD means restriction >= 52.\\nC means authorization >= 39.\\nA means condition >= 73.\\nB means insurance >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 29, \\\"authorization\\\": 37, \\\"restriction\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0013:uncertain", "group_id": "wf12:archive_loan:test:0013", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f234a1f6f9.\\nD means restriction >= 52.\\nC means authorization >= 39.\\nA means condition >= 73.\\nB means insurance >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 29, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.05, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9, \"true\": 0.1}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0014:choice", "group_id": "wf12:archive_loan:test:0014", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 95934e7aae.\\nC means authorization >= 20.\\nA means condition >= 37.\\nD means restriction >= 47.\\nB means insurance >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 13, \\\"authorization\\\": 15, \\\"restriction\\\": 97}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_184a4f\": \"Outcome PROCEED.\", \"candidate_0d4501\": \"Outcome REVIEW.\", \"candidate_1f6be6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_184a4f\": 0.0, \"candidate_0d4501\": 0.0, \"candidate_1f6be6\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0014:score", "group_id": "wf12:archive_loan:test:0014", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 95934e7aae.\\nC means authorization >= 20.\\nA means condition >= 37.\\nD means restriction >= 47.\\nB means insurance >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 13, \\\"authorization\\\": 15, \\\"restriction\\\": 97}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0014:noul", "group_id": "wf12:archive_loan:test:0014", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 95934e7aae.\\nC means authorization >= 20.\\nA means condition >= 37.\\nD means restriction >= 47.\\nB means insurance >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 13, \\\"authorization\\\": 15, \\\"restriction\\\": 97}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0014:uncertain", "group_id": "wf12:archive_loan:test:0014", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 95934e7aae.\\nC means authorization >= 20.\\nA means condition >= 37.\\nD means restriction >= 47.\\nB means insurance >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 13, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0015:choice", "group_id": "wf12:archive_loan:test:0015", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7266b949ae.\\nC means authorization >= 62.\\nB means insurance >= 50.\\nA means condition >= 26.\\nD means restriction >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 16, \\\"insurance\\\": 63, \\\"authorization\\\": 82, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bcbacb\": \"Outcome PROCEED.\", \"candidate_b57ad5\": \"Outcome REVIEW.\", \"candidate_7c5d30\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bcbacb\": 1.0, \"candidate_b57ad5\": 0.0, \"candidate_7c5d30\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0015:score", "group_id": "wf12:archive_loan:test:0015", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7266b949ae.\\nC means authorization >= 62.\\nB means insurance >= 50.\\nA means condition >= 26.\\nD means restriction >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 16, \\\"insurance\\\": 63, \\\"authorization\\\": 82, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0015:noul", "group_id": "wf12:archive_loan:test:0015", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7266b949ae.\\nC means authorization >= 62.\\nB means insurance >= 50.\\nA means condition >= 26.\\nD means restriction >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 16, \\\"insurance\\\": 63, \\\"authorization\\\": 82, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0015:uncertain", "group_id": "wf12:archive_loan:test:0015", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7266b949ae.\\nC means authorization >= 62.\\nB means insurance >= 50.\\nA means condition >= 26.\\nD means restriction >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.2, \\\"C\\\": 0.2, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bcbacb\": \"Outcome PROCEED.\", \"candidate_b57ad5\": \"Outcome REVIEW.\", \"candidate_7c5d30\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bcbacb\": 0.10873424657534247, \"candidate_b57ad5\": 0.6992657534246576, \"candidate_7c5d30\": 0.192}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0016:choice", "group_id": "wf12:archive_loan:test:0016", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fa8de1fed0.\\nA means condition >= 69.\\nC means authorization >= 20.\\nB means insurance >= 24.\\nD means restriction >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 88, \\\"insurance\\\": 66, \\\"authorization\\\": 9, \\\"restriction\\\": 43}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e8fa26\": \"Outcome PROCEED.\", \"candidate_661a3c\": \"Outcome REVIEW.\", \"candidate_623ee2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e8fa26\": 0.0, \"candidate_661a3c\": 1.0, \"candidate_623ee2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0016:score", "group_id": "wf12:archive_loan:test:0016", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fa8de1fed0.\\nA means condition >= 69.\\nC means authorization >= 20.\\nB means insurance >= 24.\\nD means restriction >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 88, \\\"insurance\\\": 66, \\\"authorization\\\": 9, \\\"restriction\\\": 43}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0016:noul", "group_id": "wf12:archive_loan:test:0016", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fa8de1fed0.\\nA means condition >= 69.\\nC means authorization >= 20.\\nB means insurance >= 24.\\nD means restriction >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 88, \\\"insurance\\\": 66, \\\"authorization\\\": 9, \\\"restriction\\\": 43}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0016:uncertain", "group_id": "wf12:archive_loan:test:0016", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fa8de1fed0.\\nA means condition >= 69.\\nC means authorization >= 20.\\nB means insurance >= 24.\\nD means restriction >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 66, \\\"authorization\\\": 9, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e8fa26\": \"Outcome PROCEED.\", \"candidate_661a3c\": \"Outcome REVIEW.\", \"candidate_623ee2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e8fa26\": 0.0, \"candidate_661a3c\": 0.050000000000000044, \"candidate_623ee2\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0017:choice", "group_id": "wf12:archive_loan:test:0017", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 1b559fda84.\\nA means condition >= 24.\\nB means insurance >= 82.\\nC means authorization >= 16.\\nD means restriction >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 2, \\\"insurance\\\": 54, \\\"authorization\\\": 11, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_07c5f7\": \"Outcome PROCEED.\", \"candidate_ab3e53\": \"Outcome REVIEW.\", \"candidate_da534c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_07c5f7\": 0.0, \"candidate_ab3e53\": 0.0, \"candidate_da534c\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0017:score", "group_id": "wf12:archive_loan:test:0017", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 1b559fda84.\\nA means condition >= 24.\\nB means insurance >= 82.\\nC means authorization >= 16.\\nD means restriction >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 2, \\\"insurance\\\": 54, \\\"authorization\\\": 11, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0017:noul", "group_id": "wf12:archive_loan:test:0017", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 1b559fda84.\\nA means condition >= 24.\\nB means insurance >= 82.\\nC means authorization >= 16.\\nD means restriction >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 2, \\\"insurance\\\": 54, \\\"authorization\\\": 11, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0017:uncertain", "group_id": "wf12:archive_loan:test:0017", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 1b559fda84.\\nA means condition >= 24.\\nB means insurance >= 82.\\nC means authorization >= 16.\\nD means restriction >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.65, \\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_07c5f7\": \"Outcome PROCEED.\", \"candidate_ab3e53\": \"Outcome REVIEW.\", \"candidate_da534c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_07c5f7\": 0.29119999999999996, \"candidate_ab3e53\": 0.43842500000000006, \"candidate_da534c\": 0.270375}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0018:choice", "group_id": "wf12:archive_loan:test:0018", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 24cd19430b.\\nC means authorization >= 38.\\nA means condition >= 71.\\nB means insurance >= 14.\\nD means restriction >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 36, \\\"insurance\\\": 79, \\\"authorization\\\": 98, \\\"restriction\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6aeffe\": \"Outcome PROCEED.\", \"candidate_75e012\": \"Outcome REVIEW.\", \"candidate_a01424\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6aeffe\": 1.0, \"candidate_75e012\": 0.0, \"candidate_a01424\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0018:score", "group_id": "wf12:archive_loan:test:0018", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 24cd19430b.\\nC means authorization >= 38.\\nA means condition >= 71.\\nB means insurance >= 14.\\nD means restriction >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 36, \\\"insurance\\\": 79, \\\"authorization\\\": 98, \\\"restriction\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0018:noul", "group_id": "wf12:archive_loan:test:0018", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 24cd19430b.\\nC means authorization >= 38.\\nA means condition >= 71.\\nB means insurance >= 14.\\nD means restriction >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 36, \\\"insurance\\\": 79, \\\"authorization\\\": 98, \\\"restriction\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0018:uncertain", "group_id": "wf12:archive_loan:test:0018", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 24cd19430b.\\nC means authorization >= 38.\\nA means condition >= 71.\\nB means insurance >= 14.\\nD means restriction >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.65, \\\"C\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.13490000000000002, \"1\": 0.16909999999999997, \"2\": 0.696}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0019:choice", "group_id": "wf12:archive_loan:test:0019", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e6288915a2.\\nC means authorization >= 20.\\nD means restriction >= 19.\\nA means condition >= 33.\\nB means insurance >= 67.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 77, \\\"insurance\\\": 98, \\\"authorization\\\": 8, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_48122f\": \"Outcome PROCEED.\", \"candidate_cc4dae\": \"Outcome REVIEW.\", \"candidate_ea5725\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_48122f\": 0.0, \"candidate_cc4dae\": 1.0, \"candidate_ea5725\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0019:score", "group_id": "wf12:archive_loan:test:0019", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e6288915a2.\\nC means authorization >= 20.\\nD means restriction >= 19.\\nA means condition >= 33.\\nB means insurance >= 67.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 77, \\\"insurance\\\": 98, \\\"authorization\\\": 8, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0019:noul", "group_id": "wf12:archive_loan:test:0019", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e6288915a2.\\nC means authorization >= 20.\\nD means restriction >= 19.\\nA means condition >= 33.\\nB means insurance >= 67.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 77, \\\"insurance\\\": 98, \\\"authorization\\\": 8, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0019:uncertain", "group_id": "wf12:archive_loan:test:0019", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e6288915a2.\\nC means authorization >= 20.\\nD means restriction >= 19.\\nA means condition >= 33.\\nB means insurance >= 67.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 77, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 10}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.05, \"1\": 0.95, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0020:choice", "group_id": "wf12:archive_loan:test:0020", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 19c0c68ace.\\nB means insurance >= 26.\\nC means authorization >= 20.\\nD means restriction >= 54.\\nA means condition >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 40, \\\"authorization\\\": 5, \\\"restriction\\\": 75}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1cd644\": \"Outcome PROCEED.\", \"candidate_76b11e\": \"Outcome REVIEW.\", \"candidate_219089\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1cd644\": 0.0, \"candidate_76b11e\": 0.0, \"candidate_219089\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0020:score", "group_id": "wf12:archive_loan:test:0020", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 19c0c68ace.\\nB means insurance >= 26.\\nC means authorization >= 20.\\nD means restriction >= 54.\\nA means condition >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 40, \\\"authorization\\\": 5, \\\"restriction\\\": 75}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0020:noul", "group_id": "wf12:archive_loan:test:0020", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 19c0c68ace.\\nB means insurance >= 26.\\nC means authorization >= 20.\\nD means restriction >= 54.\\nA means condition >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 40, \\\"authorization\\\": 5, \\\"restriction\\\": 75}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0020:uncertain", "group_id": "wf12:archive_loan:test:0020", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 19c0c68ace.\\nB means insurance >= 26.\\nC means authorization >= 20.\\nD means restriction >= 54.\\nA means condition >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 40, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8, \"1\": 0.12999999999999998, \"2\": 0.06999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0021:choice", "group_id": "wf12:archive_loan:test:0021", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7d3731c750.\\nB means insurance >= 53.\\nA means condition >= 55.\\nD means restriction >= 74.\\nC means authorization >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 69, \\\"insurance\\\": 56, \\\"authorization\\\": 41, \\\"restriction\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d883da\": \"Outcome PROCEED.\", \"candidate_bae479\": \"Outcome REVIEW.\", \"candidate_15603b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d883da\": 1.0, \"candidate_bae479\": 0.0, \"candidate_15603b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0021:score", "group_id": "wf12:archive_loan:test:0021", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7d3731c750.\\nB means insurance >= 53.\\nA means condition >= 55.\\nD means restriction >= 74.\\nC means authorization >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 69, \\\"insurance\\\": 56, \\\"authorization\\\": 41, \\\"restriction\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0021:noul", "group_id": "wf12:archive_loan:test:0021", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7d3731c750.\\nB means insurance >= 53.\\nA means condition >= 55.\\nD means restriction >= 74.\\nC means authorization >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 69, \\\"insurance\\\": 56, \\\"authorization\\\": 41, \\\"restriction\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0021:uncertain", "group_id": "wf12:archive_loan:test:0021", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7d3731c750.\\nB means insurance >= 53.\\nA means condition >= 55.\\nD means restriction >= 74.\\nC means authorization >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65, \\\"C\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.32799999999999996, \"true\": 0.672}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0022:choice", "group_id": "wf12:archive_loan:test:0022", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7e82baed9c.\\nC means authorization >= 23.\\nB means insurance >= 38.\\nA means condition >= 20.\\nD means restriction >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 41, \\\"authorization\\\": 0, \\\"restriction\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_606a7b\": \"Outcome PROCEED.\", \"candidate_d053e3\": \"Outcome REVIEW.\", \"candidate_42b965\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_606a7b\": 0.0, \"candidate_d053e3\": 1.0, \"candidate_42b965\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0022:score", "group_id": "wf12:archive_loan:test:0022", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7e82baed9c.\\nC means authorization >= 23.\\nB means insurance >= 38.\\nA means condition >= 20.\\nD means restriction >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 41, \\\"authorization\\\": 0, \\\"restriction\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0022:noul", "group_id": "wf12:archive_loan:test:0022", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7e82baed9c.\\nC means authorization >= 23.\\nB means insurance >= 38.\\nA means condition >= 20.\\nD means restriction >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 41, \\\"authorization\\\": 0, \\\"restriction\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0022:uncertain", "group_id": "wf12:archive_loan:test:0022", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7e82baed9c.\\nC means authorization >= 23.\\nB means insurance >= 38.\\nA means condition >= 20.\\nD means restriction >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.95, \\\"C\\\": 0.95, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9122499999999999, \"true\": 0.08775000000000008}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0023:choice", "group_id": "wf12:archive_loan:test:0023", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b25cefb781.\\nC means authorization >= 89.\\nB means insurance >= 90.\\nD means restriction >= 16.\\nA means condition >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 88, \\\"insurance\\\": 61, \\\"authorization\\\": 98, \\\"restriction\\\": 43}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b93c6d\": \"Outcome PROCEED.\", \"candidate_f8aab3\": \"Outcome REVIEW.\", \"candidate_03059d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b93c6d\": 0.0, \"candidate_f8aab3\": 0.0, \"candidate_03059d\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0023:score", "group_id": "wf12:archive_loan:test:0023", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b25cefb781.\\nC means authorization >= 89.\\nB means insurance >= 90.\\nD means restriction >= 16.\\nA means condition >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 88, \\\"insurance\\\": 61, \\\"authorization\\\": 98, \\\"restriction\\\": 43}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0023:noul", "group_id": "wf12:archive_loan:test:0023", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b25cefb781.\\nC means authorization >= 89.\\nB means insurance >= 90.\\nD means restriction >= 16.\\nA means condition >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 88, \\\"insurance\\\": 61, \\\"authorization\\\": 98, \\\"restriction\\\": 43}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0023:uncertain", "group_id": "wf12:archive_loan:test:0023", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b25cefb781.\\nC means authorization >= 89.\\nB means insurance >= 90.\\nD means restriction >= 16.\\nA means condition >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.35, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8245, \"true\": 0.1755}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0024:choice", "group_id": "wf12:archive_loan:test:0024", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8942840fe8.\\nA means condition >= 34.\\nD means restriction >= 62.\\nC means authorization >= 24.\\nB means insurance >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 20, \\\"authorization\\\": 73, \\\"restriction\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a3456c\": \"Outcome PROCEED.\", \"candidate_70715d\": \"Outcome REVIEW.\", \"candidate_eaa757\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a3456c\": 1.0, \"candidate_70715d\": 0.0, \"candidate_eaa757\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0024:score", "group_id": "wf12:archive_loan:test:0024", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8942840fe8.\\nA means condition >= 34.\\nD means restriction >= 62.\\nC means authorization >= 24.\\nB means insurance >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 20, \\\"authorization\\\": 73, \\\"restriction\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0024:noul", "group_id": "wf12:archive_loan:test:0024", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8942840fe8.\\nA means condition >= 34.\\nD means restriction >= 62.\\nC means authorization >= 24.\\nB means insurance >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 20, \\\"authorization\\\": 73, \\\"restriction\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0024:uncertain", "group_id": "wf12:archive_loan:test:0024", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8942840fe8.\\nA means condition >= 34.\\nD means restriction >= 62.\\nC means authorization >= 24.\\nB means insurance >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 4}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a3456c\": \"Outcome PROCEED.\", \"candidate_70715d\": \"Outcome REVIEW.\", \"candidate_eaa757\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a3456c\": 0.2, \"candidate_70715d\": 0.8, \"candidate_eaa757\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0025:choice", "group_id": "wf12:archive_loan:test:0025", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 615a67521f.\\nD means restriction >= 61.\\nC means authorization >= 24.\\nA means condition >= 66.\\nB means insurance >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 28, \\\"authorization\\\": 2, \\\"restriction\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ef7351\": \"Outcome PROCEED.\", \"candidate_346a23\": \"Outcome REVIEW.\", \"candidate_6901fd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ef7351\": 0.0, \"candidate_346a23\": 1.0, \"candidate_6901fd\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0025:score", "group_id": "wf12:archive_loan:test:0025", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 615a67521f.\\nD means restriction >= 61.\\nC means authorization >= 24.\\nA means condition >= 66.\\nB means insurance >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 28, \\\"authorization\\\": 2, \\\"restriction\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0025:noul", "group_id": "wf12:archive_loan:test:0025", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 615a67521f.\\nD means restriction >= 61.\\nC means authorization >= 24.\\nA means condition >= 66.\\nB means insurance >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 28, \\\"authorization\\\": 2, \\\"restriction\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0025:uncertain", "group_id": "wf12:archive_loan:test:0025", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 615a67521f.\\nD means restriction >= 61.\\nC means authorization >= 24.\\nA means condition >= 66.\\nB means insurance >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 28, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.1, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ef7351\": \"Outcome PROCEED.\", \"candidate_346a23\": \"Outcome REVIEW.\", \"candidate_6901fd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ef7351\": 0.095, \"candidate_346a23\": 0.855, \"candidate_6901fd\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0026:choice", "group_id": "wf12:archive_loan:test:0026", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f693e551b0.\\nB means insurance >= 73.\\nA means condition >= 89.\\nD means restriction >= 47.\\nC means authorization >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 26, \\\"insurance\\\": 71, \\\"authorization\\\": 32, \\\"restriction\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_676364\": \"Outcome PROCEED.\", \"candidate_7b49d4\": \"Outcome REVIEW.\", \"candidate_73603d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_676364\": 0.0, \"candidate_7b49d4\": 0.0, \"candidate_73603d\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0026:score", "group_id": "wf12:archive_loan:test:0026", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f693e551b0.\\nB means insurance >= 73.\\nA means condition >= 89.\\nD means restriction >= 47.\\nC means authorization >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 26, \\\"insurance\\\": 71, \\\"authorization\\\": 32, \\\"restriction\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0026:noul", "group_id": "wf12:archive_loan:test:0026", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f693e551b0.\\nB means insurance >= 73.\\nA means condition >= 89.\\nD means restriction >= 47.\\nC means authorization >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 26, \\\"insurance\\\": 71, \\\"authorization\\\": 32, \\\"restriction\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0026:uncertain", "group_id": "wf12:archive_loan:test:0026", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f693e551b0.\\nB means insurance >= 73.\\nA means condition >= 89.\\nD means restriction >= 47.\\nC means authorization >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.8, \\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_676364\": \"Outcome PROCEED.\", \"candidate_7b49d4\": \"Outcome REVIEW.\", \"candidate_73603d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_676364\": 0.7232000000000001, \"candidate_7b49d4\": 0.1508, \"candidate_73603d\": 0.12599999999999997}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0027:choice", "group_id": "wf12:archive_loan:test:0027", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference de5840a893.\\nB means insurance >= 80.\\nD means restriction >= 52.\\nC means authorization >= 52.\\nA means condition >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 14, \\\"insurance\\\": 93, \\\"authorization\\\": 68, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c9b3a1\": \"Outcome PROCEED.\", \"candidate_6040b0\": \"Outcome REVIEW.\", \"candidate_a7bad4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c9b3a1\": 1.0, \"candidate_6040b0\": 0.0, \"candidate_a7bad4\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0027:score", "group_id": "wf12:archive_loan:test:0027", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference de5840a893.\\nB means insurance >= 80.\\nD means restriction >= 52.\\nC means authorization >= 52.\\nA means condition >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 14, \\\"insurance\\\": 93, \\\"authorization\\\": 68, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0027:noul", "group_id": "wf12:archive_loan:test:0027", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference de5840a893.\\nB means insurance >= 80.\\nD means restriction >= 52.\\nC means authorization >= 52.\\nA means condition >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 14, \\\"insurance\\\": 93, \\\"authorization\\\": 68, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0027:uncertain", "group_id": "wf12:archive_loan:test:0027", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference de5840a893.\\nB means insurance >= 80.\\nD means restriction >= 52.\\nC means authorization >= 52.\\nA means condition >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 14, \\\"insurance\\\": 93, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.5369863013698629, \"1\": 0.16205479452054797, \"2\": 0.3009589041095891}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0028:choice", "group_id": "wf12:archive_loan:test:0028", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b28270c8c0.\\nC means authorization >= 86.\\nA means condition >= 20.\\nB means insurance >= 25.\\nD means restriction >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 84, \\\"authorization\\\": 61, \\\"restriction\\\": 77}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0f4c1e\": \"Outcome PROCEED.\", \"candidate_fe59ab\": \"Outcome REVIEW.\", \"candidate_4295dc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0f4c1e\": 0.0, \"candidate_fe59ab\": 1.0, \"candidate_4295dc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0028:score", "group_id": "wf12:archive_loan:test:0028", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b28270c8c0.\\nC means authorization >= 86.\\nA means condition >= 20.\\nB means insurance >= 25.\\nD means restriction >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 84, \\\"authorization\\\": 61, \\\"restriction\\\": 77}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0028:noul", "group_id": "wf12:archive_loan:test:0028", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b28270c8c0.\\nC means authorization >= 86.\\nA means condition >= 20.\\nB means insurance >= 25.\\nD means restriction >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 84, \\\"authorization\\\": 61, \\\"restriction\\\": 77}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0028:uncertain", "group_id": "wf12:archive_loan:test:0028", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b28270c8c0.\\nC means authorization >= 86.\\nA means condition >= 20.\\nB means insurance >= 25.\\nD means restriction >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 77}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0025000000000000005, \"1\": 0.9974999999999999, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0029:choice", "group_id": "wf12:archive_loan:test:0029", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 917c0b7c65.\\nD means restriction >= 11.\\nB means insurance >= 45.\\nA means condition >= 54.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 48, \\\"authorization\\\": 38, \\\"restriction\\\": 63}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_37de5c\": \"Outcome PROCEED.\", \"candidate_fa8c39\": \"Outcome REVIEW.\", \"candidate_b8ba08\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_37de5c\": 0.0, \"candidate_fa8c39\": 0.0, \"candidate_b8ba08\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0029:score", "group_id": "wf12:archive_loan:test:0029", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 917c0b7c65.\\nD means restriction >= 11.\\nB means insurance >= 45.\\nA means condition >= 54.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 48, \\\"authorization\\\": 38, \\\"restriction\\\": 63}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0029:noul", "group_id": "wf12:archive_loan:test:0029", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 917c0b7c65.\\nD means restriction >= 11.\\nB means insurance >= 45.\\nA means condition >= 54.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 48, \\\"authorization\\\": 38, \\\"restriction\\\": 63}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0029:uncertain", "group_id": "wf12:archive_loan:test:0029", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 917c0b7c65.\\nD means restriction >= 11.\\nB means insurance >= 45.\\nA means condition >= 54.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 38, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.65, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0030:choice", "group_id": "wf12:archive_loan:test:0030", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fed05a4df2.\\nD means restriction >= 39.\\nA means condition >= 29.\\nC means authorization >= 46.\\nB means insurance >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 8, \\\"insurance\\\": 100, \\\"authorization\\\": 48, \\\"restriction\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_950df1\": \"Outcome PROCEED.\", \"candidate_4b13f8\": \"Outcome REVIEW.\", \"candidate_eee56c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_950df1\": 1.0, \"candidate_4b13f8\": 0.0, \"candidate_eee56c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0030:score", "group_id": "wf12:archive_loan:test:0030", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fed05a4df2.\\nD means restriction >= 39.\\nA means condition >= 29.\\nC means authorization >= 46.\\nB means insurance >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 8, \\\"insurance\\\": 100, \\\"authorization\\\": 48, \\\"restriction\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0030:noul", "group_id": "wf12:archive_loan:test:0030", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fed05a4df2.\\nD means restriction >= 39.\\nA means condition >= 29.\\nC means authorization >= 46.\\nB means insurance >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 8, \\\"insurance\\\": 100, \\\"authorization\\\": 48, \\\"restriction\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0030:uncertain", "group_id": "wf12:archive_loan:test:0030", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference fed05a4df2.\\nD means restriction >= 39.\\nA means condition >= 29.\\nC means authorization >= 46.\\nB means insurance >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.35, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8245, \"true\": 0.17550000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0031:choice", "group_id": "wf12:archive_loan:test:0031", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6a97a44f65.\\nC means authorization >= 58.\\nB means insurance >= 46.\\nA means condition >= 10.\\nD means restriction >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 45, \\\"insurance\\\": 73, \\\"authorization\\\": 37, \\\"restriction\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_672995\": \"Outcome PROCEED.\", \"candidate_257c94\": \"Outcome REVIEW.\", \"candidate_aed075\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_672995\": 0.0, \"candidate_257c94\": 1.0, \"candidate_aed075\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0031:score", "group_id": "wf12:archive_loan:test:0031", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6a97a44f65.\\nC means authorization >= 58.\\nB means insurance >= 46.\\nA means condition >= 10.\\nD means restriction >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 45, \\\"insurance\\\": 73, \\\"authorization\\\": 37, \\\"restriction\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0031:noul", "group_id": "wf12:archive_loan:test:0031", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6a97a44f65.\\nC means authorization >= 58.\\nB means insurance >= 46.\\nA means condition >= 10.\\nD means restriction >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 45, \\\"insurance\\\": 73, \\\"authorization\\\": 37, \\\"restriction\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0031:uncertain", "group_id": "wf12:archive_loan:test:0031", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6a97a44f65.\\nC means authorization >= 58.\\nB means insurance >= 46.\\nA means condition >= 10.\\nD means restriction >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.1, \\\"C\\\": 0.05, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.054749999999999965, \"true\": 0.94525}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0032:choice", "group_id": "wf12:archive_loan:test:0032", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ebc75bbff3.\\nD means restriction >= 89.\\nB means insurance >= 57.\\nA means condition >= 88.\\nC means authorization >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 52, \\\"authorization\\\": 92, \\\"restriction\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4ec641\": \"Outcome PROCEED.\", \"candidate_e038bc\": \"Outcome REVIEW.\", \"candidate_258096\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4ec641\": 0.0, \"candidate_e038bc\": 0.0, \"candidate_258096\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0032:score", "group_id": "wf12:archive_loan:test:0032", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ebc75bbff3.\\nD means restriction >= 89.\\nB means insurance >= 57.\\nA means condition >= 88.\\nC means authorization >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 52, \\\"authorization\\\": 92, \\\"restriction\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0032:noul", "group_id": "wf12:archive_loan:test:0032", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ebc75bbff3.\\nD means restriction >= 89.\\nB means insurance >= 57.\\nA means condition >= 88.\\nC means authorization >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 52, \\\"authorization\\\": 92, \\\"restriction\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0032:uncertain", "group_id": "wf12:archive_loan:test:0032", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ebc75bbff3.\\nD means restriction >= 89.\\nB means insurance >= 57.\\nA means condition >= 88.\\nC means authorization >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.8, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.856, \"true\": 0.14400000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0033:choice", "group_id": "wf12:archive_loan:test:0033", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 411f7f4a88.\\nA means condition >= 33.\\nB means insurance >= 34.\\nC means authorization >= 18.\\nD means restriction >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 35, \\\"insurance\\\": 97, \\\"authorization\\\": 73, \\\"restriction\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_28931e\": \"Outcome PROCEED.\", \"candidate_6d2da9\": \"Outcome REVIEW.\", \"candidate_8d9180\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_28931e\": 1.0, \"candidate_6d2da9\": 0.0, \"candidate_8d9180\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0033:score", "group_id": "wf12:archive_loan:test:0033", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 411f7f4a88.\\nA means condition >= 33.\\nB means insurance >= 34.\\nC means authorization >= 18.\\nD means restriction >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 35, \\\"insurance\\\": 97, \\\"authorization\\\": 73, \\\"restriction\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0033:noul", "group_id": "wf12:archive_loan:test:0033", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 411f7f4a88.\\nA means condition >= 33.\\nB means insurance >= 34.\\nC means authorization >= 18.\\nD means restriction >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 35, \\\"insurance\\\": 97, \\\"authorization\\\": 73, \\\"restriction\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0033:uncertain", "group_id": "wf12:archive_loan:test:0033", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 411f7f4a88.\\nA means condition >= 33.\\nB means insurance >= 34.\\nC means authorization >= 18.\\nD means restriction >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 35, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 73, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_28931e\": \"Outcome PROCEED.\", \"candidate_6d2da9\": \"Outcome REVIEW.\", \"candidate_8d9180\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_28931e\": 0.9073972602739726, \"candidate_6d2da9\": 0.0, \"candidate_8d9180\": 0.09260273972602742}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0034:choice", "group_id": "wf12:archive_loan:test:0034", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b1610e2c17.\\nC means authorization >= 54.\\nA means condition >= 58.\\nD means restriction >= 45.\\nB means insurance >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 0, \\\"insurance\\\": 37, \\\"authorization\\\": 51, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_14f02f\": \"Outcome PROCEED.\", \"candidate_9f76a0\": \"Outcome REVIEW.\", \"candidate_c3332b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_14f02f\": 0.0, \"candidate_9f76a0\": 1.0, \"candidate_c3332b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0034:score", "group_id": "wf12:archive_loan:test:0034", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b1610e2c17.\\nC means authorization >= 54.\\nA means condition >= 58.\\nD means restriction >= 45.\\nB means insurance >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 0, \\\"insurance\\\": 37, \\\"authorization\\\": 51, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0034:noul", "group_id": "wf12:archive_loan:test:0034", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b1610e2c17.\\nC means authorization >= 54.\\nA means condition >= 58.\\nD means restriction >= 45.\\nB means insurance >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 0, \\\"insurance\\\": 37, \\\"authorization\\\": 51, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0034:uncertain", "group_id": "wf12:archive_loan:test:0034", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b1610e2c17.\\nC means authorization >= 54.\\nA means condition >= 58.\\nD means restriction >= 45.\\nB means insurance >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 0, \\\"insurance\\\": 37, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.95, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_14f02f\": \"Outcome PROCEED.\", \"candidate_9f76a0\": \"Outcome REVIEW.\", \"candidate_c3332b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_14f02f\": 0.95, \"candidate_9f76a0\": 0.04750000000000004, \"candidate_c3332b\": 0.0025000000000000022}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0035:choice", "group_id": "wf12:archive_loan:test:0035", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 95e6cd5bad.\\nC means authorization >= 16.\\nB means insurance >= 41.\\nA means condition >= 69.\\nD means restriction >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 15, \\\"insurance\\\": 9, \\\"authorization\\\": 10, \\\"restriction\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d449f4\": \"Outcome PROCEED.\", \"candidate_fd40cf\": \"Outcome REVIEW.\", \"candidate_8e059f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d449f4\": 0.0, \"candidate_fd40cf\": 0.0, \"candidate_8e059f\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0035:score", "group_id": "wf12:archive_loan:test:0035", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 95e6cd5bad.\\nC means authorization >= 16.\\nB means insurance >= 41.\\nA means condition >= 69.\\nD means restriction >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 15, \\\"insurance\\\": 9, \\\"authorization\\\": 10, \\\"restriction\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0035:noul", "group_id": "wf12:archive_loan:test:0035", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 95e6cd5bad.\\nC means authorization >= 16.\\nB means insurance >= 41.\\nA means condition >= 69.\\nD means restriction >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 15, \\\"insurance\\\": 9, \\\"authorization\\\": 10, \\\"restriction\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0035:uncertain", "group_id": "wf12:archive_loan:test:0035", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 95e6cd5bad.\\nC means authorization >= 16.\\nB means insurance >= 41.\\nA means condition >= 69.\\nD means restriction >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.8, \\\"C\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d449f4\": \"Outcome PROCEED.\", \"candidate_fd40cf\": \"Outcome REVIEW.\", \"candidate_8e059f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d449f4\": 0.16160000000000002, \"candidate_fd40cf\": 0.1664, \"candidate_8e059f\": 0.672}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0036:choice", "group_id": "wf12:archive_loan:test:0036", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d1097d6410.\\nC means authorization >= 15.\\nA means condition >= 23.\\nB means insurance >= 50.\\nD means restriction >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 43, \\\"insurance\\\": 88, \\\"authorization\\\": 51, \\\"restriction\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_af2222\": \"Outcome PROCEED.\", \"candidate_e06607\": \"Outcome REVIEW.\", \"candidate_90d60f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_af2222\": 1.0, \"candidate_e06607\": 0.0, \"candidate_90d60f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0036:score", "group_id": "wf12:archive_loan:test:0036", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d1097d6410.\\nC means authorization >= 15.\\nA means condition >= 23.\\nB means insurance >= 50.\\nD means restriction >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 43, \\\"insurance\\\": 88, \\\"authorization\\\": 51, \\\"restriction\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0036:noul", "group_id": "wf12:archive_loan:test:0036", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d1097d6410.\\nC means authorization >= 15.\\nA means condition >= 23.\\nB means insurance >= 50.\\nD means restriction >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 43, \\\"insurance\\\": 88, \\\"authorization\\\": 51, \\\"restriction\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0036:uncertain", "group_id": "wf12:archive_loan:test:0036", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d1097d6410.\\nC means authorization >= 15.\\nA means condition >= 23.\\nB means insurance >= 50.\\nD means restriction >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 43, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 99}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.6400000000000001, \"1\": 0.0, \"2\": 0.35999999999999993}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0037:choice", "group_id": "wf12:archive_loan:test:0037", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 27739dedbb.\\nA means condition >= 86.\\nC means authorization >= 21.\\nD means restriction >= 12.\\nB means insurance >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 36, \\\"authorization\\\": 15, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ed477f\": \"Outcome PROCEED.\", \"candidate_692cb9\": \"Outcome REVIEW.\", \"candidate_3bbdbc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ed477f\": 0.0, \"candidate_692cb9\": 1.0, \"candidate_3bbdbc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0037:score", "group_id": "wf12:archive_loan:test:0037", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 27739dedbb.\\nA means condition >= 86.\\nC means authorization >= 21.\\nD means restriction >= 12.\\nB means insurance >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 36, \\\"authorization\\\": 15, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0037:noul", "group_id": "wf12:archive_loan:test:0037", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 27739dedbb.\\nA means condition >= 86.\\nC means authorization >= 21.\\nD means restriction >= 12.\\nB means insurance >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 36, \\\"authorization\\\": 15, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0037:uncertain", "group_id": "wf12:archive_loan:test:0037", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 27739dedbb.\\nA means condition >= 86.\\nC means authorization >= 21.\\nD means restriction >= 12.\\nB means insurance >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 8}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"C\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.855, \"1\": 0.14500000000000002, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0038:choice", "group_id": "wf12:archive_loan:test:0038", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2ddc0b06c0.\\nB means insurance >= 87.\\nD means restriction >= 50.\\nC means authorization >= 87.\\nA means condition >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 90, \\\"insurance\\\": 57, \\\"authorization\\\": 98, \\\"restriction\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_95f6a4\": \"Outcome PROCEED.\", \"candidate_150cd9\": \"Outcome REVIEW.\", \"candidate_d3b8e6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_95f6a4\": 0.0, \"candidate_150cd9\": 0.0, \"candidate_d3b8e6\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0038:score", "group_id": "wf12:archive_loan:test:0038", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2ddc0b06c0.\\nB means insurance >= 87.\\nD means restriction >= 50.\\nC means authorization >= 87.\\nA means condition >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 90, \\\"insurance\\\": 57, \\\"authorization\\\": 98, \\\"restriction\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0038:noul", "group_id": "wf12:archive_loan:test:0038", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2ddc0b06c0.\\nB means insurance >= 87.\\nD means restriction >= 50.\\nC means authorization >= 87.\\nA means condition >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 90, \\\"insurance\\\": 57, \\\"authorization\\\": 98, \\\"restriction\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0038:uncertain", "group_id": "wf12:archive_loan:test:0038", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2ddc0b06c0.\\nB means insurance >= 87.\\nD means restriction >= 50.\\nC means authorization >= 87.\\nA means condition >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 57, \\\"authorization\\\": 98, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.06999999999999998, \"1\": 0.12999999999999998, \"2\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0039:choice", "group_id": "wf12:archive_loan:test:0039", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9837dcf8d5.\\nB means insurance >= 78.\\nD means restriction >= 10.\\nC means authorization >= 23.\\nA means condition >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 88, \\\"authorization\\\": 51, \\\"restriction\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d26634\": \"Outcome PROCEED.\", \"candidate_005de8\": \"Outcome REVIEW.\", \"candidate_522404\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d26634\": 1.0, \"candidate_005de8\": 0.0, \"candidate_522404\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0039:score", "group_id": "wf12:archive_loan:test:0039", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9837dcf8d5.\\nB means insurance >= 78.\\nD means restriction >= 10.\\nC means authorization >= 23.\\nA means condition >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 88, \\\"authorization\\\": 51, \\\"restriction\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0039:noul", "group_id": "wf12:archive_loan:test:0039", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9837dcf8d5.\\nB means insurance >= 78.\\nD means restriction >= 10.\\nC means authorization >= 23.\\nA means condition >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 88, \\\"authorization\\\": 51, \\\"restriction\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0039:uncertain", "group_id": "wf12:archive_loan:test:0039", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9837dcf8d5.\\nB means insurance >= 78.\\nD means restriction >= 10.\\nC means authorization >= 23.\\nA means condition >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.8, \\\"C\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.706, \"true\": 0.294}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0040:choice", "group_id": "wf12:archive_loan:test:0040", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3161cbcaee.\\nA means condition >= 18.\\nC means authorization >= 64.\\nD means restriction >= 60.\\nB means insurance >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 78, \\\"authorization\\\": 19, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a1c431\": \"Outcome PROCEED.\", \"candidate_97d6fc\": \"Outcome REVIEW.\", \"candidate_8410f8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a1c431\": 0.0, \"candidate_97d6fc\": 1.0, \"candidate_8410f8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0040:score", "group_id": "wf12:archive_loan:test:0040", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3161cbcaee.\\nA means condition >= 18.\\nC means authorization >= 64.\\nD means restriction >= 60.\\nB means insurance >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 78, \\\"authorization\\\": 19, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0040:noul", "group_id": "wf12:archive_loan:test:0040", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3161cbcaee.\\nA means condition >= 18.\\nC means authorization >= 64.\\nD means restriction >= 60.\\nB means insurance >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 78, \\\"authorization\\\": 19, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0040:uncertain", "group_id": "wf12:archive_loan:test:0040", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3161cbcaee.\\nA means condition >= 18.\\nC means authorization >= 64.\\nD means restriction >= 60.\\nB means insurance >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.1, \\\"C\\\": 0.1, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9505, \"true\": 0.04950000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0041:choice", "group_id": "wf12:archive_loan:test:0041", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 310be92481.\\nA means condition >= 57.\\nB means insurance >= 61.\\nD means restriction >= 82.\\nC means authorization >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 5, \\\"authorization\\\": 53, \\\"restriction\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_346cce\": \"Outcome PROCEED.\", \"candidate_602c66\": \"Outcome REVIEW.\", \"candidate_a2afc3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_346cce\": 0.0, \"candidate_602c66\": 0.0, \"candidate_a2afc3\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0041:score", "group_id": "wf12:archive_loan:test:0041", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 310be92481.\\nA means condition >= 57.\\nB means insurance >= 61.\\nD means restriction >= 82.\\nC means authorization >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 5, \\\"authorization\\\": 53, \\\"restriction\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0041:noul", "group_id": "wf12:archive_loan:test:0041", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 310be92481.\\nA means condition >= 57.\\nB means insurance >= 61.\\nD means restriction >= 82.\\nC means authorization >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 5, \\\"authorization\\\": 53, \\\"restriction\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0041:uncertain", "group_id": "wf12:archive_loan:test:0041", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 310be92481.\\nA means condition >= 57.\\nB means insurance >= 61.\\nD means restriction >= 82.\\nC means authorization >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 5, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0042:choice", "group_id": "wf12:archive_loan:test:0042", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 10abd3faca.\\nA means condition >= 53.\\nB means insurance >= 35.\\nD means restriction >= 86.\\nC means authorization >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 23, \\\"insurance\\\": 61, \\\"authorization\\\": 41, \\\"restriction\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_98ea68\": \"Outcome PROCEED.\", \"candidate_f300e8\": \"Outcome REVIEW.\", \"candidate_986fc0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_98ea68\": 1.0, \"candidate_f300e8\": 0.0, \"candidate_986fc0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0042:score", "group_id": "wf12:archive_loan:test:0042", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 10abd3faca.\\nA means condition >= 53.\\nB means insurance >= 35.\\nD means restriction >= 86.\\nC means authorization >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 23, \\\"insurance\\\": 61, \\\"authorization\\\": 41, \\\"restriction\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0042:noul", "group_id": "wf12:archive_loan:test:0042", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 10abd3faca.\\nA means condition >= 53.\\nB means insurance >= 35.\\nD means restriction >= 86.\\nC means authorization >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 23, \\\"insurance\\\": 61, \\\"authorization\\\": 41, \\\"restriction\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0042:uncertain", "group_id": "wf12:archive_loan:test:0042", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 10abd3faca.\\nA means condition >= 53.\\nB means insurance >= 35.\\nD means restriction >= 86.\\nC means authorization >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 23, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 41, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_98ea68\": \"Outcome PROCEED.\", \"candidate_f300e8\": \"Outcome REVIEW.\", \"candidate_986fc0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_98ea68\": 0.8000000000000002, \"candidate_f300e8\": 0.15999999999999998, \"candidate_986fc0\": 0.039999999999999994}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0043:choice", "group_id": "wf12:archive_loan:test:0043", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 69397db108.\\nB means insurance >= 83.\\nA means condition >= 42.\\nC means authorization >= 23.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 6, \\\"insurance\\\": 56, \\\"authorization\\\": 63, \\\"restriction\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_97f986\": \"Outcome PROCEED.\", \"candidate_b2fa05\": \"Outcome REVIEW.\", \"candidate_fe6d11\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_97f986\": 0.0, \"candidate_b2fa05\": 1.0, \"candidate_fe6d11\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0043:score", "group_id": "wf12:archive_loan:test:0043", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 69397db108.\\nB means insurance >= 83.\\nA means condition >= 42.\\nC means authorization >= 23.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 6, \\\"insurance\\\": 56, \\\"authorization\\\": 63, \\\"restriction\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0043:noul", "group_id": "wf12:archive_loan:test:0043", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 69397db108.\\nB means insurance >= 83.\\nA means condition >= 42.\\nC means authorization >= 23.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 6, \\\"insurance\\\": 56, \\\"authorization\\\": 63, \\\"restriction\\\": 9}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0043:uncertain", "group_id": "wf12:archive_loan:test:0043", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 69397db108.\\nB means insurance >= 83.\\nA means condition >= 42.\\nC means authorization >= 23.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 6, \\\"insurance\\\": 56, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.1, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_97f986\": \"Outcome PROCEED.\", \"candidate_b2fa05\": \"Outcome REVIEW.\", \"candidate_fe6d11\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_97f986\": 0.0, \"candidate_b2fa05\": 0.95, \"candidate_fe6d11\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0044:choice", "group_id": "wf12:archive_loan:test:0044", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 69033e83e7.\\nC means authorization >= 78.\\nB means insurance >= 10.\\nA means condition >= 80.\\nD means restriction >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 66, \\\"insurance\\\": 3, \\\"authorization\\\": 24, \\\"restriction\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8c8849\": \"Outcome PROCEED.\", \"candidate_f79370\": \"Outcome REVIEW.\", \"candidate_0a534a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8c8849\": 0.0, \"candidate_f79370\": 0.0, \"candidate_0a534a\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0044:score", "group_id": "wf12:archive_loan:test:0044", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 69033e83e7.\\nC means authorization >= 78.\\nB means insurance >= 10.\\nA means condition >= 80.\\nD means restriction >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 66, \\\"insurance\\\": 3, \\\"authorization\\\": 24, \\\"restriction\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0044:noul", "group_id": "wf12:archive_loan:test:0044", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 69033e83e7.\\nC means authorization >= 78.\\nB means insurance >= 10.\\nA means condition >= 80.\\nD means restriction >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 66, \\\"insurance\\\": 3, \\\"authorization\\\": 24, \\\"restriction\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0044:uncertain", "group_id": "wf12:archive_loan:test:0044", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 69033e83e7.\\nC means authorization >= 78.\\nB means insurance >= 10.\\nA means condition >= 80.\\nD means restriction >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 66, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 67}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8c8849\": \"Outcome PROCEED.\", \"candidate_f79370\": \"Outcome REVIEW.\", \"candidate_0a534a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8c8849\": 0.12249999999999998, \"candidate_f79370\": 0.0, \"candidate_0a534a\": 0.8775}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0045:choice", "group_id": "wf12:archive_loan:test:0045", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f3a10d1e36.\\nC means authorization >= 30.\\nD means restriction >= 66.\\nB means insurance >= 83.\\nA means condition >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 98, \\\"insurance\\\": 87, \\\"authorization\\\": 91, \\\"restriction\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c3be85\": \"Outcome PROCEED.\", \"candidate_fef82c\": \"Outcome REVIEW.\", \"candidate_94290a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c3be85\": 1.0, \"candidate_fef82c\": 0.0, \"candidate_94290a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0045:score", "group_id": "wf12:archive_loan:test:0045", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f3a10d1e36.\\nC means authorization >= 30.\\nD means restriction >= 66.\\nB means insurance >= 83.\\nA means condition >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 98, \\\"insurance\\\": 87, \\\"authorization\\\": 91, \\\"restriction\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0045:noul", "group_id": "wf12:archive_loan:test:0045", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f3a10d1e36.\\nC means authorization >= 30.\\nD means restriction >= 66.\\nB means insurance >= 83.\\nA means condition >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 98, \\\"insurance\\\": 87, \\\"authorization\\\": 91, \\\"restriction\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0045:uncertain", "group_id": "wf12:archive_loan:test:0045", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f3a10d1e36.\\nC means authorization >= 30.\\nD means restriction >= 66.\\nB means insurance >= 83.\\nA means condition >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 87, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 94}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.0, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0046:choice", "group_id": "wf12:archive_loan:test:0046", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f265d89907.\\nD means restriction >= 44.\\nA means condition >= 79.\\nB means insurance >= 80.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 65, \\\"insurance\\\": 15, \\\"authorization\\\": 36, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d1d18c\": \"Outcome PROCEED.\", \"candidate_4b0310\": \"Outcome REVIEW.\", \"candidate_df5d87\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d1d18c\": 0.0, \"candidate_4b0310\": 1.0, \"candidate_df5d87\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0046:score", "group_id": "wf12:archive_loan:test:0046", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f265d89907.\\nD means restriction >= 44.\\nA means condition >= 79.\\nB means insurance >= 80.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 65, \\\"insurance\\\": 15, \\\"authorization\\\": 36, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0046:noul", "group_id": "wf12:archive_loan:test:0046", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f265d89907.\\nD means restriction >= 44.\\nA means condition >= 79.\\nB means insurance >= 80.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 65, \\\"insurance\\\": 15, \\\"authorization\\\": 36, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0046:uncertain", "group_id": "wf12:archive_loan:test:0046", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f265d89907.\\nD means restriction >= 44.\\nA means condition >= 79.\\nB means insurance >= 80.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 65, \\\"insurance\\\": 15, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.95, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.09999999999999998, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0047:choice", "group_id": "wf12:archive_loan:test:0047", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 940bdb3056.\\nA means condition >= 12.\\nD means restriction >= 37.\\nB means insurance >= 39.\\nC means authorization >= 83.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 34, \\\"insurance\\\": 27, \\\"authorization\\\": 70, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d31ed0\": \"Outcome PROCEED.\", \"candidate_57d259\": \"Outcome REVIEW.\", \"candidate_6eb759\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d31ed0\": 0.0, \"candidate_57d259\": 0.0, \"candidate_6eb759\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0047:score", "group_id": "wf12:archive_loan:test:0047", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 940bdb3056.\\nA means condition >= 12.\\nD means restriction >= 37.\\nB means insurance >= 39.\\nC means authorization >= 83.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 34, \\\"insurance\\\": 27, \\\"authorization\\\": 70, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0047:noul", "group_id": "wf12:archive_loan:test:0047", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 940bdb3056.\\nA means condition >= 12.\\nD means restriction >= 37.\\nB means insurance >= 39.\\nC means authorization >= 83.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 34, \\\"insurance\\\": 27, \\\"authorization\\\": 70, \\\"restriction\\\": 39}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0047:uncertain", "group_id": "wf12:archive_loan:test:0047", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 940bdb3056.\\nA means condition >= 12.\\nD means restriction >= 37.\\nB means insurance >= 39.\\nC means authorization >= 83.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 34, \\\"insurance\\\": 27, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.22749999999999998, \"1\": 0.12249999999999998, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0048:choice", "group_id": "wf12:archive_loan:test:0048", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4a381bca6f.\\nC means authorization >= 65.\\nD means restriction >= 27.\\nA means condition >= 29.\\nB means insurance >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 29, \\\"insurance\\\": 96, \\\"authorization\\\": 69, \\\"restriction\\\": 92}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_711889\": \"Outcome PROCEED.\", \"candidate_0c64fd\": \"Outcome REVIEW.\", \"candidate_23f1fd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_711889\": 1.0, \"candidate_0c64fd\": 0.0, \"candidate_23f1fd\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0048:score", "group_id": "wf12:archive_loan:test:0048", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4a381bca6f.\\nC means authorization >= 65.\\nD means restriction >= 27.\\nA means condition >= 29.\\nB means insurance >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 29, \\\"insurance\\\": 96, \\\"authorization\\\": 69, \\\"restriction\\\": 92}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0048:noul", "group_id": "wf12:archive_loan:test:0048", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4a381bca6f.\\nC means authorization >= 65.\\nD means restriction >= 27.\\nA means condition >= 29.\\nB means insurance >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 29, \\\"insurance\\\": 96, \\\"authorization\\\": 69, \\\"restriction\\\": 92}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0048:uncertain", "group_id": "wf12:archive_loan:test:0048", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4a381bca6f.\\nC means authorization >= 65.\\nD means restriction >= 27.\\nA means condition >= 29.\\nB means insurance >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 29, \\\"insurance\\\": 96, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.12999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0049:choice", "group_id": "wf12:archive_loan:test:0049", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 05bca726ce.\\nB means insurance >= 57.\\nD means restriction >= 71.\\nC means authorization >= 84.\\nA means condition >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 43, \\\"authorization\\\": 98, \\\"restriction\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9b959a\": \"Outcome PROCEED.\", \"candidate_e5b1ef\": \"Outcome REVIEW.\", \"candidate_0e6360\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9b959a\": 0.0, \"candidate_e5b1ef\": 1.0, \"candidate_0e6360\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0049:score", "group_id": "wf12:archive_loan:test:0049", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 05bca726ce.\\nB means insurance >= 57.\\nD means restriction >= 71.\\nC means authorization >= 84.\\nA means condition >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 43, \\\"authorization\\\": 98, \\\"restriction\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0049:noul", "group_id": "wf12:archive_loan:test:0049", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 05bca726ce.\\nB means insurance >= 57.\\nD means restriction >= 71.\\nC means authorization >= 84.\\nA means condition >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 43, \\\"authorization\\\": 98, \\\"restriction\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0049:uncertain", "group_id": "wf12:archive_loan:test:0049", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 05bca726ce.\\nB means insurance >= 57.\\nD means restriction >= 71.\\nC means authorization >= 84.\\nA means condition >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.95, \\\"C\\\": 0.05, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.14274999999999993, \"true\": 0.8572500000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0050:choice", "group_id": "wf12:archive_loan:test:0050", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8c3b310299.\\nC means authorization >= 17.\\nB means insurance >= 87.\\nA means condition >= 87.\\nD means restriction >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 76, \\\"insurance\\\": 1, \\\"authorization\\\": 28, \\\"restriction\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ef4710\": \"Outcome PROCEED.\", \"candidate_c0a43f\": \"Outcome REVIEW.\", \"candidate_df828a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ef4710\": 0.0, \"candidate_c0a43f\": 0.0, \"candidate_df828a\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0050:score", "group_id": "wf12:archive_loan:test:0050", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8c3b310299.\\nC means authorization >= 17.\\nB means insurance >= 87.\\nA means condition >= 87.\\nD means restriction >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 76, \\\"insurance\\\": 1, \\\"authorization\\\": 28, \\\"restriction\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0050:noul", "group_id": "wf12:archive_loan:test:0050", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8c3b310299.\\nC means authorization >= 17.\\nB means insurance >= 87.\\nA means condition >= 87.\\nD means restriction >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 76, \\\"insurance\\\": 1, \\\"authorization\\\": 28, \\\"restriction\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0050:uncertain", "group_id": "wf12:archive_loan:test:0050", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8c3b310299.\\nC means authorization >= 17.\\nB means insurance >= 87.\\nA means condition >= 87.\\nD means restriction >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 28, \\\"restriction\\\": 82}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.65, \"true\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0051:choice", "group_id": "wf12:archive_loan:test:0051", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 24431a9215.\\nA means condition >= 37.\\nD means restriction >= 40.\\nB means insurance >= 35.\\nC means authorization >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 91, \\\"insurance\\\": 33, \\\"authorization\\\": 90, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eedaef\": \"Outcome PROCEED.\", \"candidate_832079\": \"Outcome REVIEW.\", \"candidate_4a8fab\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eedaef\": 1.0, \"candidate_832079\": 0.0, \"candidate_4a8fab\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0051:score", "group_id": "wf12:archive_loan:test:0051", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 24431a9215.\\nA means condition >= 37.\\nD means restriction >= 40.\\nB means insurance >= 35.\\nC means authorization >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 91, \\\"insurance\\\": 33, \\\"authorization\\\": 90, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0051:noul", "group_id": "wf12:archive_loan:test:0051", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 24431a9215.\\nA means condition >= 37.\\nD means restriction >= 40.\\nB means insurance >= 35.\\nC means authorization >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 91, \\\"insurance\\\": 33, \\\"authorization\\\": 90, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0051:uncertain", "group_id": "wf12:archive_loan:test:0051", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 24431a9215.\\nA means condition >= 37.\\nD means restriction >= 40.\\nB means insurance >= 35.\\nC means authorization >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 33, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 14}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eedaef\": \"Outcome PROCEED.\", \"candidate_832079\": \"Outcome REVIEW.\", \"candidate_4a8fab\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eedaef\": 0.27999999999999997, \"candidate_832079\": 0.72, \"candidate_4a8fab\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0052:choice", "group_id": "wf12:archive_loan:test:0052", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference de6ce95911.\\nC means authorization >= 53.\\nD means restriction >= 11.\\nA means condition >= 11.\\nB means insurance >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 66, \\\"insurance\\\": 92, \\\"authorization\\\": 21, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cb5297\": \"Outcome PROCEED.\", \"candidate_d7c6c4\": \"Outcome REVIEW.\", \"candidate_f9c998\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cb5297\": 0.0, \"candidate_d7c6c4\": 1.0, \"candidate_f9c998\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0052:score", "group_id": "wf12:archive_loan:test:0052", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference de6ce95911.\\nC means authorization >= 53.\\nD means restriction >= 11.\\nA means condition >= 11.\\nB means insurance >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 66, \\\"insurance\\\": 92, \\\"authorization\\\": 21, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0052:noul", "group_id": "wf12:archive_loan:test:0052", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference de6ce95911.\\nC means authorization >= 53.\\nD means restriction >= 11.\\nA means condition >= 11.\\nB means insurance >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 66, \\\"insurance\\\": 92, \\\"authorization\\\": 21, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0052:uncertain", "group_id": "wf12:archive_loan:test:0052", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference de6ce95911.\\nC means authorization >= 53.\\nD means restriction >= 11.\\nA means condition >= 11.\\nB means insurance >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 92, \\\"authorization\\\": 21, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cb5297\": \"Outcome PROCEED.\", \"candidate_d7c6c4\": \"Outcome REVIEW.\", \"candidate_f9c998\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cb5297\": 0.0, \"candidate_d7c6c4\": 0.95, \"candidate_f9c998\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0053:choice", "group_id": "wf12:archive_loan:test:0053", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3d59c30d2c.\\nA means condition >= 68.\\nD means restriction >= 83.\\nC means authorization >= 80.\\nB means insurance >= 36.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 56, \\\"insurance\\\": 59, \\\"authorization\\\": 53, \\\"restriction\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0b97c6\": \"Outcome PROCEED.\", \"candidate_bad625\": \"Outcome REVIEW.\", \"candidate_922872\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0b97c6\": 0.0, \"candidate_bad625\": 0.0, \"candidate_922872\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0053:score", "group_id": "wf12:archive_loan:test:0053", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3d59c30d2c.\\nA means condition >= 68.\\nD means restriction >= 83.\\nC means authorization >= 80.\\nB means insurance >= 36.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 56, \\\"insurance\\\": 59, \\\"authorization\\\": 53, \\\"restriction\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0053:noul", "group_id": "wf12:archive_loan:test:0053", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3d59c30d2c.\\nA means condition >= 68.\\nD means restriction >= 83.\\nC means authorization >= 80.\\nB means insurance >= 36.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 56, \\\"insurance\\\": 59, \\\"authorization\\\": 53, \\\"restriction\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0053:uncertain", "group_id": "wf12:archive_loan:test:0053", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3d59c30d2c.\\nA means condition >= 68.\\nD means restriction >= 83.\\nC means authorization >= 80.\\nB means insurance >= 36.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 56, \\\"insurance\\\": 59, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0b97c6\": \"Outcome PROCEED.\", \"candidate_bad625\": \"Outcome REVIEW.\", \"candidate_922872\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0b97c6\": 0.35, \"candidate_bad625\": 0.12999999999999998, \"candidate_922872\": 0.52}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0054:choice", "group_id": "wf12:archive_loan:test:0054", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 889c96f836.\\nB means insurance >= 63.\\nC means authorization >= 58.\\nA means condition >= 65.\\nD means restriction >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": 86, \\\"authorization\\\": 89, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d1511c\": \"Outcome PROCEED.\", \"candidate_07a0c4\": \"Outcome REVIEW.\", \"candidate_0f514a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d1511c\": 1.0, \"candidate_07a0c4\": 0.0, \"candidate_0f514a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0054:score", "group_id": "wf12:archive_loan:test:0054", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 889c96f836.\\nB means insurance >= 63.\\nC means authorization >= 58.\\nA means condition >= 65.\\nD means restriction >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": 86, \\\"authorization\\\": 89, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0054:noul", "group_id": "wf12:archive_loan:test:0054", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 889c96f836.\\nB means insurance >= 63.\\nC means authorization >= 58.\\nA means condition >= 65.\\nD means restriction >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": 86, \\\"authorization\\\": 89, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0054:uncertain", "group_id": "wf12:archive_loan:test:0054", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 889c96f836.\\nB means insurance >= 63.\\nC means authorization >= 58.\\nA means condition >= 65.\\nD means restriction >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 89, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8960000000000001, \"1\": 0.0831999999999999, \"2\": 0.020799999999999975}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0055:choice", "group_id": "wf12:archive_loan:test:0055", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d318ce8064.\\nB means insurance >= 24.\\nA means condition >= 60.\\nC means authorization >= 14.\\nD means restriction >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": 13, \\\"authorization\\\": 8, \\\"restriction\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_850b9e\": \"Outcome PROCEED.\", \"candidate_1c9793\": \"Outcome REVIEW.\", \"candidate_ccf0e9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_850b9e\": 0.0, \"candidate_1c9793\": 1.0, \"candidate_ccf0e9\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0055:score", "group_id": "wf12:archive_loan:test:0055", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d318ce8064.\\nB means insurance >= 24.\\nA means condition >= 60.\\nC means authorization >= 14.\\nD means restriction >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": 13, \\\"authorization\\\": 8, \\\"restriction\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0055:noul", "group_id": "wf12:archive_loan:test:0055", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d318ce8064.\\nB means insurance >= 24.\\nA means condition >= 60.\\nC means authorization >= 14.\\nD means restriction >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": 13, \\\"authorization\\\": 8, \\\"restriction\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0055:uncertain", "group_id": "wf12:archive_loan:test:0055", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d318ce8064.\\nB means insurance >= 24.\\nA means condition >= 60.\\nC means authorization >= 14.\\nD means restriction >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 8, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.09999999999999998, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0056:choice", "group_id": "wf12:archive_loan:test:0056", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0031104d4a.\\nA means condition >= 59.\\nC means authorization >= 28.\\nD means restriction >= 75.\\nB means insurance >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 70, \\\"insurance\\\": 40, \\\"authorization\\\": 8, \\\"restriction\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e2d6c2\": \"Outcome PROCEED.\", \"candidate_169d03\": \"Outcome REVIEW.\", \"candidate_d3e63e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e2d6c2\": 0.0, \"candidate_169d03\": 0.0, \"candidate_d3e63e\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0056:score", "group_id": "wf12:archive_loan:test:0056", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0031104d4a.\\nA means condition >= 59.\\nC means authorization >= 28.\\nD means restriction >= 75.\\nB means insurance >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 70, \\\"insurance\\\": 40, \\\"authorization\\\": 8, \\\"restriction\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0056:noul", "group_id": "wf12:archive_loan:test:0056", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0031104d4a.\\nA means condition >= 59.\\nC means authorization >= 28.\\nD means restriction >= 75.\\nB means insurance >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 70, \\\"insurance\\\": 40, \\\"authorization\\\": 8, \\\"restriction\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0056:uncertain", "group_id": "wf12:archive_loan:test:0056", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0031104d4a.\\nA means condition >= 59.\\nC means authorization >= 28.\\nD means restriction >= 75.\\nB means insurance >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 70, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 8, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.19999999999999996, \"2\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0057:choice", "group_id": "wf12:archive_loan:test:0057", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 48e8b81a83.\\nA means condition >= 12.\\nB means insurance >= 88.\\nC means authorization >= 14.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 74, \\\"insurance\\\": 95, \\\"authorization\\\": 46, \\\"restriction\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8f5084\": \"Outcome PROCEED.\", \"candidate_ba1314\": \"Outcome REVIEW.\", \"candidate_4b3071\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8f5084\": 1.0, \"candidate_ba1314\": 0.0, \"candidate_4b3071\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0057:score", "group_id": "wf12:archive_loan:test:0057", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 48e8b81a83.\\nA means condition >= 12.\\nB means insurance >= 88.\\nC means authorization >= 14.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 74, \\\"insurance\\\": 95, \\\"authorization\\\": 46, \\\"restriction\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0057:noul", "group_id": "wf12:archive_loan:test:0057", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 48e8b81a83.\\nA means condition >= 12.\\nB means insurance >= 88.\\nC means authorization >= 14.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 74, \\\"insurance\\\": 95, \\\"authorization\\\": 46, \\\"restriction\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0057:uncertain", "group_id": "wf12:archive_loan:test:0057", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 48e8b81a83.\\nA means condition >= 12.\\nB means insurance >= 88.\\nC means authorization >= 14.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 46, \\\"restriction\\\": 90}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.65, \"true\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0058:choice", "group_id": "wf12:archive_loan:test:0058", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f7f20bd865.\\nA means condition >= 69.\\nD means restriction >= 40.\\nC means authorization >= 89.\\nB means insurance >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 35, \\\"insurance\\\": 25, \\\"authorization\\\": 19, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_977c9d\": \"Outcome PROCEED.\", \"candidate_bc94b1\": \"Outcome REVIEW.\", \"candidate_aab9b3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_977c9d\": 0.0, \"candidate_bc94b1\": 1.0, \"candidate_aab9b3\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0058:score", "group_id": "wf12:archive_loan:test:0058", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f7f20bd865.\\nA means condition >= 69.\\nD means restriction >= 40.\\nC means authorization >= 89.\\nB means insurance >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 35, \\\"insurance\\\": 25, \\\"authorization\\\": 19, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0058:noul", "group_id": "wf12:archive_loan:test:0058", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f7f20bd865.\\nA means condition >= 69.\\nD means restriction >= 40.\\nC means authorization >= 89.\\nB means insurance >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 35, \\\"insurance\\\": 25, \\\"authorization\\\": 19, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0058:uncertain", "group_id": "wf12:archive_loan:test:0058", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f7f20bd865.\\nA means condition >= 69.\\nD means restriction >= 40.\\nC means authorization >= 89.\\nB means insurance >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 35, \\\"insurance\\\": 25, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.95, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.95, \"true\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0059:choice", "group_id": "wf12:archive_loan:test:0059", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b1744ae111.\\nD means restriction >= 50.\\nC means authorization >= 51.\\nA means condition >= 10.\\nB means insurance >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 44, \\\"authorization\\\": 63, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_94d0bf\": \"Outcome PROCEED.\", \"candidate_d82383\": \"Outcome REVIEW.\", \"candidate_3e9b73\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_94d0bf\": 0.0, \"candidate_d82383\": 0.0, \"candidate_3e9b73\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0059:score", "group_id": "wf12:archive_loan:test:0059", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b1744ae111.\\nD means restriction >= 50.\\nC means authorization >= 51.\\nA means condition >= 10.\\nB means insurance >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 44, \\\"authorization\\\": 63, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0059:noul", "group_id": "wf12:archive_loan:test:0059", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b1744ae111.\\nD means restriction >= 50.\\nC means authorization >= 51.\\nA means condition >= 10.\\nB means insurance >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 44, \\\"authorization\\\": 63, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0059:uncertain", "group_id": "wf12:archive_loan:test:0059", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b1744ae111.\\nD means restriction >= 50.\\nC means authorization >= 51.\\nA means condition >= 10.\\nB means insurance >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 44, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.19999999999999996, \"true\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0060:choice", "group_id": "wf12:archive_loan:test:0060", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 684171249e.\\nC means authorization >= 82.\\nD means restriction >= 74.\\nB means insurance >= 20.\\nA means condition >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 40, \\\"authorization\\\": 100, \\\"restriction\\\": 64}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_49d45c\": \"Outcome PROCEED.\", \"candidate_2e4834\": \"Outcome REVIEW.\", \"candidate_ad7e07\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_49d45c\": 1.0, \"candidate_2e4834\": 0.0, \"candidate_ad7e07\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0060:score", "group_id": "wf12:archive_loan:test:0060", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 684171249e.\\nC means authorization >= 82.\\nD means restriction >= 74.\\nB means insurance >= 20.\\nA means condition >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 40, \\\"authorization\\\": 100, \\\"restriction\\\": 64}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0060:noul", "group_id": "wf12:archive_loan:test:0060", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 684171249e.\\nC means authorization >= 82.\\nD means restriction >= 74.\\nB means insurance >= 20.\\nA means condition >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 40, \\\"authorization\\\": 100, \\\"restriction\\\": 64}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0060:uncertain", "group_id": "wf12:archive_loan:test:0060", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 684171249e.\\nC means authorization >= 82.\\nD means restriction >= 74.\\nB means insurance >= 20.\\nA means condition >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65, \\\"C\\\": 0.35, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_49d45c\": \"Outcome PROCEED.\", \"candidate_2e4834\": \"Outcome REVIEW.\", \"candidate_ad7e07\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_49d45c\": 0.2471, \"candidate_2e4834\": 0.13489999999999996, \"candidate_ad7e07\": 0.6180000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0061:choice", "group_id": "wf12:archive_loan:test:0061", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 49aff5161f.\\nD means restriction >= 46.\\nB means insurance >= 12.\\nC means authorization >= 89.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 9, \\\"authorization\\\": 59, \\\"restriction\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ed4117\": \"Outcome PROCEED.\", \"candidate_799cbe\": \"Outcome REVIEW.\", \"candidate_539708\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ed4117\": 0.0, \"candidate_799cbe\": 1.0, \"candidate_539708\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0061:score", "group_id": "wf12:archive_loan:test:0061", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 49aff5161f.\\nD means restriction >= 46.\\nB means insurance >= 12.\\nC means authorization >= 89.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 9, \\\"authorization\\\": 59, \\\"restriction\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0061:noul", "group_id": "wf12:archive_loan:test:0061", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 49aff5161f.\\nD means restriction >= 46.\\nB means insurance >= 12.\\nC means authorization >= 89.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 9, \\\"authorization\\\": 59, \\\"restriction\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0061:uncertain", "group_id": "wf12:archive_loan:test:0061", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 49aff5161f.\\nD means restriction >= 46.\\nB means insurance >= 12.\\nC means authorization >= 89.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 9, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.05, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ed4117\": \"Outcome PROCEED.\", \"candidate_799cbe\": \"Outcome REVIEW.\", \"candidate_539708\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ed4117\": 0.0, \"candidate_799cbe\": 0.9, \"candidate_539708\": 0.1}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0062:choice", "group_id": "wf12:archive_loan:test:0062", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 491ed30011.\\nC means authorization >= 16.\\nA means condition >= 57.\\nD means restriction >= 36.\\nB means insurance >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 71, \\\"insurance\\\": 29, \\\"authorization\\\": 6, \\\"restriction\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_19ca98\": \"Outcome PROCEED.\", \"candidate_6391bf\": \"Outcome REVIEW.\", \"candidate_7162f9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_19ca98\": 0.0, \"candidate_6391bf\": 0.0, \"candidate_7162f9\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0062:score", "group_id": "wf12:archive_loan:test:0062", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 491ed30011.\\nC means authorization >= 16.\\nA means condition >= 57.\\nD means restriction >= 36.\\nB means insurance >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 71, \\\"insurance\\\": 29, \\\"authorization\\\": 6, \\\"restriction\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0062:noul", "group_id": "wf12:archive_loan:test:0062", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 491ed30011.\\nC means authorization >= 16.\\nA means condition >= 57.\\nD means restriction >= 36.\\nB means insurance >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 71, \\\"insurance\\\": 29, \\\"authorization\\\": 6, \\\"restriction\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0062:uncertain", "group_id": "wf12:archive_loan:test:0062", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 491ed30011.\\nC means authorization >= 16.\\nA means condition >= 57.\\nD means restriction >= 36.\\nB means insurance >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 71, \\\"insurance\\\": 29, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_19ca98\": \"Outcome PROCEED.\", \"candidate_6391bf\": \"Outcome REVIEW.\", \"candidate_7162f9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_19ca98\": 0.27999999999999997, \"candidate_6391bf\": 0.06999999999999998, \"candidate_7162f9\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0063:choice", "group_id": "wf12:archive_loan:test:0063", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c1042c3057.\\nB means insurance >= 20.\\nA means condition >= 38.\\nC means authorization >= 38.\\nD means restriction >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": 100, \\\"authorization\\\": 94, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1eb842\": \"Outcome PROCEED.\", \"candidate_b1d417\": \"Outcome REVIEW.\", \"candidate_5646a4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1eb842\": 1.0, \"candidate_b1d417\": 0.0, \"candidate_5646a4\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0063:score", "group_id": "wf12:archive_loan:test:0063", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c1042c3057.\\nB means insurance >= 20.\\nA means condition >= 38.\\nC means authorization >= 38.\\nD means restriction >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": 100, \\\"authorization\\\": 94, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0063:noul", "group_id": "wf12:archive_loan:test:0063", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c1042c3057.\\nB means insurance >= 20.\\nA means condition >= 38.\\nC means authorization >= 38.\\nD means restriction >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": 100, \\\"authorization\\\": 94, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0063:uncertain", "group_id": "wf12:archive_loan:test:0063", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c1042c3057.\\nB means insurance >= 20.\\nA means condition >= 38.\\nC means authorization >= 38.\\nD means restriction >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 30}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.6400000000000001, \"1\": 0.35999999999999993, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0064:choice", "group_id": "wf12:archive_loan:test:0064", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5225f0940d.\\nA means condition >= 60.\\nD means restriction >= 78.\\nB means insurance >= 68.\\nC means authorization >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 63, \\\"insurance\\\": 85, \\\"authorization\\\": 24, \\\"restriction\\\": 19}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a8786f\": \"Outcome PROCEED.\", \"candidate_a2b49f\": \"Outcome REVIEW.\", \"candidate_31a812\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a8786f\": 0.0, \"candidate_a2b49f\": 1.0, \"candidate_31a812\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0064:score", "group_id": "wf12:archive_loan:test:0064", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5225f0940d.\\nA means condition >= 60.\\nD means restriction >= 78.\\nB means insurance >= 68.\\nC means authorization >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 63, \\\"insurance\\\": 85, \\\"authorization\\\": 24, \\\"restriction\\\": 19}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0064:noul", "group_id": "wf12:archive_loan:test:0064", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5225f0940d.\\nA means condition >= 60.\\nD means restriction >= 78.\\nB means insurance >= 68.\\nC means authorization >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 63, \\\"insurance\\\": 85, \\\"authorization\\\": 24, \\\"restriction\\\": 19}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0064:uncertain", "group_id": "wf12:archive_loan:test:0064", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5225f0940d.\\nA means condition >= 60.\\nD means restriction >= 78.\\nB means insurance >= 68.\\nC means authorization >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 63, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 24, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.09999999999999998, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0065:choice", "group_id": "wf12:archive_loan:test:0065", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d1e23dae86.\\nC means authorization >= 82.\\nA means condition >= 34.\\nD means restriction >= 85.\\nB means insurance >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": 10, \\\"authorization\\\": 56, \\\"restriction\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e55cea\": \"Outcome PROCEED.\", \"candidate_1df177\": \"Outcome REVIEW.\", \"candidate_8a310c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e55cea\": 0.0, \"candidate_1df177\": 0.0, \"candidate_8a310c\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0065:score", "group_id": "wf12:archive_loan:test:0065", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d1e23dae86.\\nC means authorization >= 82.\\nA means condition >= 34.\\nD means restriction >= 85.\\nB means insurance >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": 10, \\\"authorization\\\": 56, \\\"restriction\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0065:noul", "group_id": "wf12:archive_loan:test:0065", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d1e23dae86.\\nC means authorization >= 82.\\nA means condition >= 34.\\nD means restriction >= 85.\\nB means insurance >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": 10, \\\"authorization\\\": 56, \\\"restriction\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0065:uncertain", "group_id": "wf12:archive_loan:test:0065", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d1e23dae86.\\nC means authorization >= 82.\\nA means condition >= 34.\\nD means restriction >= 85.\\nB means insurance >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 10, \\\"authorization\\\": 56, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.8, \"2\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0066:choice", "group_id": "wf12:archive_loan:test:0066", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 07d6a911da.\\nB means insurance >= 49.\\nC means authorization >= 71.\\nD means restriction >= 12.\\nA means condition >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 51, \\\"authorization\\\": 99, \\\"restriction\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eaf963\": \"Outcome PROCEED.\", \"candidate_4e4055\": \"Outcome REVIEW.\", \"candidate_c35b4a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eaf963\": 1.0, \"candidate_4e4055\": 0.0, \"candidate_c35b4a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0066:score", "group_id": "wf12:archive_loan:test:0066", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 07d6a911da.\\nB means insurance >= 49.\\nC means authorization >= 71.\\nD means restriction >= 12.\\nA means condition >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 51, \\\"authorization\\\": 99, \\\"restriction\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0066:noul", "group_id": "wf12:archive_loan:test:0066", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 07d6a911da.\\nB means insurance >= 49.\\nC means authorization >= 71.\\nD means restriction >= 12.\\nA means condition >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 81, \\\"insurance\\\": 51, \\\"authorization\\\": 99, \\\"restriction\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0066:uncertain", "group_id": "wf12:archive_loan:test:0066", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 07d6a911da.\\nB means insurance >= 49.\\nC means authorization >= 71.\\nD means restriction >= 12.\\nA means condition >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65, \\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.748, \"true\": 0.252}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0067:choice", "group_id": "wf12:archive_loan:test:0067", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3dc8a0cb9b.\\nC means authorization >= 47.\\nA means condition >= 21.\\nB means insurance >= 79.\\nD means restriction >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": 42, \\\"authorization\\\": 40, \\\"restriction\\\": 44}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3f6bd3\": \"Outcome PROCEED.\", \"candidate_5e7e9b\": \"Outcome REVIEW.\", \"candidate_dda045\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3f6bd3\": 0.0, \"candidate_5e7e9b\": 1.0, \"candidate_dda045\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0067:score", "group_id": "wf12:archive_loan:test:0067", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3dc8a0cb9b.\\nC means authorization >= 47.\\nA means condition >= 21.\\nB means insurance >= 79.\\nD means restriction >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": 42, \\\"authorization\\\": 40, \\\"restriction\\\": 44}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0067:noul", "group_id": "wf12:archive_loan:test:0067", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3dc8a0cb9b.\\nC means authorization >= 47.\\nA means condition >= 21.\\nB means insurance >= 79.\\nD means restriction >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": 42, \\\"authorization\\\": 40, \\\"restriction\\\": 44}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0067:uncertain", "group_id": "wf12:archive_loan:test:0067", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3dc8a0cb9b.\\nC means authorization >= 47.\\nA means condition >= 21.\\nB means insurance >= 79.\\nD means restriction >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.95, \\\"C\\\": 0.95, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.99025, \"true\": 0.009750000000000009}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0068:choice", "group_id": "wf12:archive_loan:test:0068", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c133c0427c.\\nC means authorization >= 59.\\nD means restriction >= 44.\\nA means condition >= 77.\\nB means insurance >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 2, \\\"authorization\\\": 87, \\\"restriction\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_429408\": \"Outcome PROCEED.\", \"candidate_6af026\": \"Outcome REVIEW.\", \"candidate_98ac94\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_429408\": 0.0, \"candidate_6af026\": 0.0, \"candidate_98ac94\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0068:score", "group_id": "wf12:archive_loan:test:0068", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c133c0427c.\\nC means authorization >= 59.\\nD means restriction >= 44.\\nA means condition >= 77.\\nB means insurance >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 2, \\\"authorization\\\": 87, \\\"restriction\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0068:noul", "group_id": "wf12:archive_loan:test:0068", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c133c0427c.\\nC means authorization >= 59.\\nD means restriction >= 44.\\nA means condition >= 77.\\nB means insurance >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 2, \\\"authorization\\\": 87, \\\"restriction\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0068:uncertain", "group_id": "wf12:archive_loan:test:0068", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c133c0427c.\\nC means authorization >= 59.\\nD means restriction >= 44.\\nA means condition >= 77.\\nB means insurance >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 79}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"C\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.4225000000000001, \"true\": 0.5774999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0069:choice", "group_id": "wf12:archive_loan:test:0069", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cf6acb901e.\\nD means restriction >= 52.\\nA means condition >= 85.\\nB means insurance >= 52.\\nC means authorization >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 87, \\\"insurance\\\": 7, \\\"authorization\\\": 82, \\\"restriction\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6ca211\": \"Outcome PROCEED.\", \"candidate_643cf9\": \"Outcome REVIEW.\", \"candidate_4e2587\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6ca211\": 1.0, \"candidate_643cf9\": 0.0, \"candidate_4e2587\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0069:score", "group_id": "wf12:archive_loan:test:0069", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cf6acb901e.\\nD means restriction >= 52.\\nA means condition >= 85.\\nB means insurance >= 52.\\nC means authorization >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 87, \\\"insurance\\\": 7, \\\"authorization\\\": 82, \\\"restriction\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0069:noul", "group_id": "wf12:archive_loan:test:0069", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cf6acb901e.\\nD means restriction >= 52.\\nA means condition >= 85.\\nB means insurance >= 52.\\nC means authorization >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 87, \\\"insurance\\\": 7, \\\"authorization\\\": 82, \\\"restriction\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0069:uncertain", "group_id": "wf12:archive_loan:test:0069", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cf6acb901e.\\nD means restriction >= 52.\\nA means condition >= 85.\\nB means insurance >= 52.\\nC means authorization >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 87, \\\"insurance\\\": 7, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6ca211\": \"Outcome PROCEED.\", \"candidate_643cf9\": \"Outcome REVIEW.\", \"candidate_4e2587\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6ca211\": 0.42958904109589036, \"candidate_643cf9\": 0.3704109589041097, \"candidate_4e2587\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0070:choice", "group_id": "wf12:archive_loan:test:0070", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9cac395613.\\nA means condition >= 76.\\nD means restriction >= 12.\\nB means insurance >= 10.\\nC means authorization >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 88, \\\"authorization\\\": 9, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9f4b5d\": \"Outcome PROCEED.\", \"candidate_dd5688\": \"Outcome REVIEW.\", \"candidate_153612\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9f4b5d\": 0.0, \"candidate_dd5688\": 1.0, \"candidate_153612\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0070:score", "group_id": "wf12:archive_loan:test:0070", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9cac395613.\\nA means condition >= 76.\\nD means restriction >= 12.\\nB means insurance >= 10.\\nC means authorization >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 88, \\\"authorization\\\": 9, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0070:noul", "group_id": "wf12:archive_loan:test:0070", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9cac395613.\\nA means condition >= 76.\\nD means restriction >= 12.\\nB means insurance >= 10.\\nC means authorization >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 88, \\\"authorization\\\": 9, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0070:uncertain", "group_id": "wf12:archive_loan:test:0070", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9cac395613.\\nA means condition >= 76.\\nD means restriction >= 12.\\nB means insurance >= 10.\\nC means authorization >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 88, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 5}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9f4b5d\": \"Outcome PROCEED.\", \"candidate_dd5688\": \"Outcome REVIEW.\", \"candidate_153612\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9f4b5d\": 0.95, \"candidate_dd5688\": 0.050000000000000044, \"candidate_153612\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0071:choice", "group_id": "wf12:archive_loan:test:0071", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 616b5a73ee.\\nA means condition >= 48.\\nD means restriction >= 90.\\nC means authorization >= 16.\\nB means insurance >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 64, \\\"insurance\\\": 79, \\\"authorization\\\": 7, \\\"restriction\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ca51dd\": \"Outcome PROCEED.\", \"candidate_1f4d9c\": \"Outcome REVIEW.\", \"candidate_49a7cd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ca51dd\": 0.0, \"candidate_1f4d9c\": 0.0, \"candidate_49a7cd\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0071:score", "group_id": "wf12:archive_loan:test:0071", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 616b5a73ee.\\nA means condition >= 48.\\nD means restriction >= 90.\\nC means authorization >= 16.\\nB means insurance >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 64, \\\"insurance\\\": 79, \\\"authorization\\\": 7, \\\"restriction\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0071:noul", "group_id": "wf12:archive_loan:test:0071", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 616b5a73ee.\\nA means condition >= 48.\\nD means restriction >= 90.\\nC means authorization >= 16.\\nB means insurance >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 64, \\\"insurance\\\": 79, \\\"authorization\\\": 7, \\\"restriction\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0071:uncertain", "group_id": "wf12:archive_loan:test:0071", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 616b5a73ee.\\nA means condition >= 48.\\nD means restriction >= 90.\\nC means authorization >= 16.\\nB means insurance >= 25.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.35, \\\"C\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ca51dd\": \"Outcome PROCEED.\", \"candidate_1f4d9c\": \"Outcome REVIEW.\", \"candidate_49a7cd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ca51dd\": 0.25707499999999994, \"candidate_1f4d9c\": 0.12492499999999998, \"candidate_49a7cd\": 0.618}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0072:choice", "group_id": "wf12:archive_loan:test:0072", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b0e4fd0ea2.\\nA means condition >= 82.\\nD means restriction >= 64.\\nC means authorization >= 60.\\nB means insurance >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 27, \\\"authorization\\\": 97, \\\"restriction\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ee9832\": \"Outcome PROCEED.\", \"candidate_16506a\": \"Outcome REVIEW.\", \"candidate_d22a38\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ee9832\": 1.0, \"candidate_16506a\": 0.0, \"candidate_d22a38\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0072:score", "group_id": "wf12:archive_loan:test:0072", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b0e4fd0ea2.\\nA means condition >= 82.\\nD means restriction >= 64.\\nC means authorization >= 60.\\nB means insurance >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 27, \\\"authorization\\\": 97, \\\"restriction\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0072:noul", "group_id": "wf12:archive_loan:test:0072", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b0e4fd0ea2.\\nA means condition >= 82.\\nD means restriction >= 64.\\nC means authorization >= 60.\\nB means insurance >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": 27, \\\"authorization\\\": 97, \\\"restriction\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0072:uncertain", "group_id": "wf12:archive_loan:test:0072", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b0e4fd0ea2.\\nA means condition >= 82.\\nD means restriction >= 64.\\nC means authorization >= 60.\\nB means insurance >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 84, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 97, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9299999999999999, \"1\": 0.0, \"2\": 0.06999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0073:choice", "group_id": "wf12:archive_loan:test:0073", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 1c2105918e.\\nC means authorization >= 57.\\nD means restriction >= 26.\\nA means condition >= 19.\\nB means insurance >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": 9, \\\"authorization\\\": 96, \\\"restriction\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1ed341\": \"Outcome PROCEED.\", \"candidate_41a61d\": \"Outcome REVIEW.\", \"candidate_fc3050\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1ed341\": 0.0, \"candidate_41a61d\": 1.0, \"candidate_fc3050\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0073:score", "group_id": "wf12:archive_loan:test:0073", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 1c2105918e.\\nC means authorization >= 57.\\nD means restriction >= 26.\\nA means condition >= 19.\\nB means insurance >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": 9, \\\"authorization\\\": 96, \\\"restriction\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0073:noul", "group_id": "wf12:archive_loan:test:0073", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 1c2105918e.\\nC means authorization >= 57.\\nD means restriction >= 26.\\nA means condition >= 19.\\nB means insurance >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": 9, \\\"authorization\\\": 96, \\\"restriction\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0073:uncertain", "group_id": "wf12:archive_loan:test:0073", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 1c2105918e.\\nC means authorization >= 57.\\nD means restriction >= 26.\\nA means condition >= 19.\\nB means insurance >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 96, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.1, \"1\": 0.08999999999999998, \"2\": 0.81}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0074:choice", "group_id": "wf12:archive_loan:test:0074", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0f3c545ebd.\\nA means condition >= 69.\\nD means restriction >= 34.\\nB means insurance >= 38.\\nC means authorization >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 39, \\\"insurance\\\": 20, \\\"authorization\\\": 96, \\\"restriction\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b2d23d\": \"Outcome PROCEED.\", \"candidate_c45f35\": \"Outcome REVIEW.\", \"candidate_a1a919\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b2d23d\": 0.0, \"candidate_c45f35\": 0.0, \"candidate_a1a919\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0074:score", "group_id": "wf12:archive_loan:test:0074", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0f3c545ebd.\\nA means condition >= 69.\\nD means restriction >= 34.\\nB means insurance >= 38.\\nC means authorization >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 39, \\\"insurance\\\": 20, \\\"authorization\\\": 96, \\\"restriction\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0074:noul", "group_id": "wf12:archive_loan:test:0074", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0f3c545ebd.\\nA means condition >= 69.\\nD means restriction >= 34.\\nB means insurance >= 38.\\nC means authorization >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 39, \\\"insurance\\\": 20, \\\"authorization\\\": 96, \\\"restriction\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0074:uncertain", "group_id": "wf12:archive_loan:test:0074", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0f3c545ebd.\\nA means condition >= 69.\\nD means restriction >= 34.\\nB means insurance >= 38.\\nC means authorization >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 96, \\\"restriction\\\": 67}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.0, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0075:choice", "group_id": "wf12:archive_loan:test:0075", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 59ff06ebf5.\\nD means restriction >= 13.\\nA means condition >= 14.\\nC means authorization >= 70.\\nB means insurance >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 11, \\\"authorization\\\": 89, \\\"restriction\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c7153c\": \"Outcome PROCEED.\", \"candidate_b622b4\": \"Outcome REVIEW.\", \"candidate_cf3413\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c7153c\": 1.0, \"candidate_b622b4\": 0.0, \"candidate_cf3413\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0075:score", "group_id": "wf12:archive_loan:test:0075", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 59ff06ebf5.\\nD means restriction >= 13.\\nA means condition >= 14.\\nC means authorization >= 70.\\nB means insurance >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 11, \\\"authorization\\\": 89, \\\"restriction\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0075:noul", "group_id": "wf12:archive_loan:test:0075", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 59ff06ebf5.\\nD means restriction >= 13.\\nA means condition >= 14.\\nC means authorization >= 70.\\nB means insurance >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 11, \\\"authorization\\\": 89, \\\"restriction\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0075:uncertain", "group_id": "wf12:archive_loan:test:0075", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 59ff06ebf5.\\nD means restriction >= 13.\\nA means condition >= 14.\\nC means authorization >= 70.\\nB means insurance >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 11, \\\"authorization\\\": 89, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0076:choice", "group_id": "wf12:archive_loan:test:0076", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 965a13dba5.\\nB means insurance >= 81.\\nD means restriction >= 50.\\nA means condition >= 17.\\nC means authorization >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": 38, \\\"authorization\\\": 10, \\\"restriction\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_75e3a5\": \"Outcome PROCEED.\", \"candidate_8092a2\": \"Outcome REVIEW.\", \"candidate_33db28\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_75e3a5\": 0.0, \"candidate_8092a2\": 1.0, \"candidate_33db28\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0076:score", "group_id": "wf12:archive_loan:test:0076", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 965a13dba5.\\nB means insurance >= 81.\\nD means restriction >= 50.\\nA means condition >= 17.\\nC means authorization >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": 38, \\\"authorization\\\": 10, \\\"restriction\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0076:noul", "group_id": "wf12:archive_loan:test:0076", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 965a13dba5.\\nB means insurance >= 81.\\nD means restriction >= 50.\\nA means condition >= 17.\\nC means authorization >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": 38, \\\"authorization\\\": 10, \\\"restriction\\\": 49}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0076:uncertain", "group_id": "wf12:archive_loan:test:0076", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 965a13dba5.\\nB means insurance >= 81.\\nD means restriction >= 50.\\nA means condition >= 17.\\nC means authorization >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.1, \\\"C\\\": 0.05, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.054750000000000076, \"true\": 0.9452499999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0077:choice", "group_id": "wf12:archive_loan:test:0077", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b5633158b2.\\nA means condition >= 13.\\nD means restriction >= 64.\\nC means authorization >= 23.\\nB means insurance >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 17, \\\"authorization\\\": 75, \\\"restriction\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_248ba9\": \"Outcome PROCEED.\", \"candidate_6bfbf1\": \"Outcome REVIEW.\", \"candidate_706de8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_248ba9\": 0.0, \"candidate_6bfbf1\": 0.0, \"candidate_706de8\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0077:score", "group_id": "wf12:archive_loan:test:0077", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b5633158b2.\\nA means condition >= 13.\\nD means restriction >= 64.\\nC means authorization >= 23.\\nB means insurance >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 17, \\\"authorization\\\": 75, \\\"restriction\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0077:noul", "group_id": "wf12:archive_loan:test:0077", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b5633158b2.\\nA means condition >= 13.\\nD means restriction >= 64.\\nC means authorization >= 23.\\nB means insurance >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 17, \\\"authorization\\\": 75, \\\"restriction\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0077:uncertain", "group_id": "wf12:archive_loan:test:0077", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b5633158b2.\\nA means condition >= 13.\\nD means restriction >= 64.\\nC means authorization >= 23.\\nB means insurance >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.35, \\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.692875, \"true\": 0.307125}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0078:choice", "group_id": "wf12:archive_loan:test:0078", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e98ae4d6e6.\\nD means restriction >= 25.\\nA means condition >= 78.\\nB means insurance >= 85.\\nC means authorization >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": 94, \\\"authorization\\\": 95, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cc6a54\": \"Outcome PROCEED.\", \"candidate_0d98a2\": \"Outcome REVIEW.\", \"candidate_d7a643\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cc6a54\": 1.0, \"candidate_0d98a2\": 0.0, \"candidate_d7a643\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0078:score", "group_id": "wf12:archive_loan:test:0078", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e98ae4d6e6.\\nD means restriction >= 25.\\nA means condition >= 78.\\nB means insurance >= 85.\\nC means authorization >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": 94, \\\"authorization\\\": 95, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0078:noul", "group_id": "wf12:archive_loan:test:0078", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e98ae4d6e6.\\nD means restriction >= 25.\\nA means condition >= 78.\\nB means insurance >= 85.\\nC means authorization >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": 94, \\\"authorization\\\": 95, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0078:uncertain", "group_id": "wf12:archive_loan:test:0078", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e98ae4d6e6.\\nD means restriction >= 25.\\nA means condition >= 78.\\nB means insurance >= 85.\\nC means authorization >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 22, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 10}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cc6a54\": \"Outcome PROCEED.\", \"candidate_0d98a2\": \"Outcome REVIEW.\", \"candidate_d7a643\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cc6a54\": 0.27999999999999997, \"candidate_0d98a2\": 0.72, \"candidate_d7a643\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0079:choice", "group_id": "wf12:archive_loan:test:0079", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b3074fb22b.\\nC means authorization >= 48.\\nA means condition >= 64.\\nB means insurance >= 73.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 53, \\\"insurance\\\": 62, \\\"authorization\\\": 32, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7a1c7d\": \"Outcome PROCEED.\", \"candidate_5bd142\": \"Outcome REVIEW.\", \"candidate_dcfebe\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7a1c7d\": 0.0, \"candidate_5bd142\": 1.0, \"candidate_dcfebe\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0079:score", "group_id": "wf12:archive_loan:test:0079", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b3074fb22b.\\nC means authorization >= 48.\\nA means condition >= 64.\\nB means insurance >= 73.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 53, \\\"insurance\\\": 62, \\\"authorization\\\": 32, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0079:noul", "group_id": "wf12:archive_loan:test:0079", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b3074fb22b.\\nC means authorization >= 48.\\nA means condition >= 64.\\nB means insurance >= 73.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 53, \\\"insurance\\\": 62, \\\"authorization\\\": 32, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0079:uncertain", "group_id": "wf12:archive_loan:test:0079", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b3074fb22b.\\nC means authorization >= 48.\\nA means condition >= 64.\\nB means insurance >= 73.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 53, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 32, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7a1c7d\": \"Outcome PROCEED.\", \"candidate_5bd142\": \"Outcome REVIEW.\", \"candidate_dcfebe\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7a1c7d\": 0.0, \"candidate_5bd142\": 0.9, \"candidate_dcfebe\": 0.1}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0080:choice", "group_id": "wf12:archive_loan:test:0080", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 580aab34b4.\\nC means authorization >= 70.\\nB means insurance >= 83.\\nD means restriction >= 79.\\nA means condition >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 80, \\\"insurance\\\": 39, \\\"authorization\\\": 75, \\\"restriction\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4630eb\": \"Outcome PROCEED.\", \"candidate_079f8b\": \"Outcome REVIEW.\", \"candidate_783e0b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4630eb\": 0.0, \"candidate_079f8b\": 0.0, \"candidate_783e0b\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0080:score", "group_id": "wf12:archive_loan:test:0080", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 580aab34b4.\\nC means authorization >= 70.\\nB means insurance >= 83.\\nD means restriction >= 79.\\nA means condition >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 80, \\\"insurance\\\": 39, \\\"authorization\\\": 75, \\\"restriction\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0080:noul", "group_id": "wf12:archive_loan:test:0080", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 580aab34b4.\\nC means authorization >= 70.\\nB means insurance >= 83.\\nD means restriction >= 79.\\nA means condition >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 80, \\\"insurance\\\": 39, \\\"authorization\\\": 75, \\\"restriction\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0080:uncertain", "group_id": "wf12:archive_loan:test:0080", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 580aab34b4.\\nC means authorization >= 70.\\nB means insurance >= 83.\\nD means restriction >= 79.\\nA means condition >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 75, \\\"restriction\\\": 96}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4630eb\": \"Outcome PROCEED.\", \"candidate_079f8b\": \"Outcome REVIEW.\", \"candidate_783e0b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4630eb\": 0.8000000000000002, \"candidate_079f8b\": 0.0, \"candidate_783e0b\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0081:choice", "group_id": "wf12:archive_loan:test:0081", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2c1abebfbc.\\nD means restriction >= 60.\\nC means authorization >= 85.\\nB means insurance >= 59.\\nA means condition >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 61, \\\"insurance\\\": 94, \\\"authorization\\\": 95, \\\"restriction\\\": 86}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0feb1d\": \"Outcome PROCEED.\", \"candidate_d15afb\": \"Outcome REVIEW.\", \"candidate_28c608\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0feb1d\": 1.0, \"candidate_d15afb\": 0.0, \"candidate_28c608\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0081:score", "group_id": "wf12:archive_loan:test:0081", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2c1abebfbc.\\nD means restriction >= 60.\\nC means authorization >= 85.\\nB means insurance >= 59.\\nA means condition >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 61, \\\"insurance\\\": 94, \\\"authorization\\\": 95, \\\"restriction\\\": 86}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0081:noul", "group_id": "wf12:archive_loan:test:0081", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2c1abebfbc.\\nD means restriction >= 60.\\nC means authorization >= 85.\\nB means insurance >= 59.\\nA means condition >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 61, \\\"insurance\\\": 94, \\\"authorization\\\": 95, \\\"restriction\\\": 86}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0081:uncertain", "group_id": "wf12:archive_loan:test:0081", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2c1abebfbc.\\nD means restriction >= 60.\\nC means authorization >= 85.\\nB means insurance >= 59.\\nA means condition >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 61, \\\"insurance\\\": 94, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8960000000000001, \"1\": 0.020799999999999968, \"2\": 0.0831999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0082:choice", "group_id": "wf12:archive_loan:test:0082", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a4a6e62cbf.\\nA means condition >= 68.\\nD means restriction >= 82.\\nC means authorization >= 21.\\nB means insurance >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 89, \\\"insurance\\\": 10, \\\"authorization\\\": 9, \\\"restriction\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_52a1d8\": \"Outcome PROCEED.\", \"candidate_ca9cb4\": \"Outcome REVIEW.\", \"candidate_0610e6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_52a1d8\": 0.0, \"candidate_ca9cb4\": 1.0, \"candidate_0610e6\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0082:score", "group_id": "wf12:archive_loan:test:0082", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a4a6e62cbf.\\nA means condition >= 68.\\nD means restriction >= 82.\\nC means authorization >= 21.\\nB means insurance >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 89, \\\"insurance\\\": 10, \\\"authorization\\\": 9, \\\"restriction\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0082:noul", "group_id": "wf12:archive_loan:test:0082", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a4a6e62cbf.\\nA means condition >= 68.\\nD means restriction >= 82.\\nC means authorization >= 21.\\nB means insurance >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 89, \\\"insurance\\\": 10, \\\"authorization\\\": 9, \\\"restriction\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0082:uncertain", "group_id": "wf12:archive_loan:test:0082", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a4a6e62cbf.\\nA means condition >= 68.\\nD means restriction >= 82.\\nC means authorization >= 21.\\nB means insurance >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 10, \\\"authorization\\\": 9, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.9, \"2\": 0.1}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0083:choice", "group_id": "wf12:archive_loan:test:0083", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8459c619ae.\\nA means condition >= 27.\\nD means restriction >= 72.\\nB means insurance >= 43.\\nC means authorization >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 98, \\\"insurance\\\": 1, \\\"authorization\\\": 73, \\\"restriction\\\": 75}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_37b9f6\": \"Outcome PROCEED.\", \"candidate_a7cc94\": \"Outcome REVIEW.\", \"candidate_3d9b50\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_37b9f6\": 0.0, \"candidate_a7cc94\": 0.0, \"candidate_3d9b50\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0083:score", "group_id": "wf12:archive_loan:test:0083", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8459c619ae.\\nA means condition >= 27.\\nD means restriction >= 72.\\nB means insurance >= 43.\\nC means authorization >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 98, \\\"insurance\\\": 1, \\\"authorization\\\": 73, \\\"restriction\\\": 75}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0083:noul", "group_id": "wf12:archive_loan:test:0083", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8459c619ae.\\nA means condition >= 27.\\nD means restriction >= 72.\\nB means insurance >= 43.\\nC means authorization >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 98, \\\"insurance\\\": 1, \\\"authorization\\\": 73, \\\"restriction\\\": 75}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0083:uncertain", "group_id": "wf12:archive_loan:test:0083", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8459c619ae.\\nA means condition >= 27.\\nD means restriction >= 72.\\nB means insurance >= 43.\\nC means authorization >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.65, \\\"C\\\": 0.2, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.14120000000000002, \"1\": 0.6848000000000001, \"2\": 0.17400000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0084:choice", "group_id": "wf12:archive_loan:test:0084", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e2d36c0cd3.\\nD means restriction >= 32.\\nB means insurance >= 20.\\nC means authorization >= 15.\\nA means condition >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 49, \\\"authorization\\\": 94, \\\"restriction\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f0ba07\": \"Outcome PROCEED.\", \"candidate_f76ed2\": \"Outcome REVIEW.\", \"candidate_c4bc52\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f0ba07\": 1.0, \"candidate_f76ed2\": 0.0, \"candidate_c4bc52\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0084:score", "group_id": "wf12:archive_loan:test:0084", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e2d36c0cd3.\\nD means restriction >= 32.\\nB means insurance >= 20.\\nC means authorization >= 15.\\nA means condition >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 49, \\\"authorization\\\": 94, \\\"restriction\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0084:noul", "group_id": "wf12:archive_loan:test:0084", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e2d36c0cd3.\\nD means restriction >= 32.\\nB means insurance >= 20.\\nC means authorization >= 15.\\nA means condition >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 49, \\\"authorization\\\": 94, \\\"restriction\\\": 74}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0084:uncertain", "group_id": "wf12:archive_loan:test:0084", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e2d36c0cd3.\\nD means restriction >= 32.\\nB means insurance >= 20.\\nC means authorization >= 15.\\nA means condition >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 49, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.699041095890411, \"true\": 0.3009589041095891}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0085:choice", "group_id": "wf12:archive_loan:test:0085", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ddc6a5029b.\\nB means insurance >= 57.\\nD means restriction >= 35.\\nA means condition >= 56.\\nC means authorization >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 73, \\\"insurance\\\": 36, \\\"authorization\\\": 3, \\\"restriction\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3e9ec2\": \"Outcome PROCEED.\", \"candidate_baf0eb\": \"Outcome REVIEW.\", \"candidate_92be69\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3e9ec2\": 0.0, \"candidate_baf0eb\": 1.0, \"candidate_92be69\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0085:score", "group_id": "wf12:archive_loan:test:0085", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ddc6a5029b.\\nB means insurance >= 57.\\nD means restriction >= 35.\\nA means condition >= 56.\\nC means authorization >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 73, \\\"insurance\\\": 36, \\\"authorization\\\": 3, \\\"restriction\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0085:noul", "group_id": "wf12:archive_loan:test:0085", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ddc6a5029b.\\nB means insurance >= 57.\\nD means restriction >= 35.\\nA means condition >= 56.\\nC means authorization >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 73, \\\"insurance\\\": 36, \\\"authorization\\\": 3, \\\"restriction\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0085:uncertain", "group_id": "wf12:archive_loan:test:0085", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ddc6a5029b.\\nB means insurance >= 57.\\nD means restriction >= 35.\\nA means condition >= 56.\\nC means authorization >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.95, \\\"C\\\": 0.1, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.1854999999999999, \"true\": 0.8145000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0086:choice", "group_id": "wf12:archive_loan:test:0086", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0759e75a1c.\\nB means insurance >= 53.\\nC means authorization >= 59.\\nD means restriction >= 48.\\nA means condition >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 32, \\\"insurance\\\": 20, \\\"authorization\\\": 55, \\\"restriction\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5d1907\": \"Outcome PROCEED.\", \"candidate_9ac2fd\": \"Outcome REVIEW.\", \"candidate_2a5a26\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5d1907\": 0.0, \"candidate_9ac2fd\": 0.0, \"candidate_2a5a26\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0086:score", "group_id": "wf12:archive_loan:test:0086", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0759e75a1c.\\nB means insurance >= 53.\\nC means authorization >= 59.\\nD means restriction >= 48.\\nA means condition >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 32, \\\"insurance\\\": 20, \\\"authorization\\\": 55, \\\"restriction\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0086:noul", "group_id": "wf12:archive_loan:test:0086", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0759e75a1c.\\nB means insurance >= 53.\\nC means authorization >= 59.\\nD means restriction >= 48.\\nA means condition >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 32, \\\"insurance\\\": 20, \\\"authorization\\\": 55, \\\"restriction\\\": 72}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0086:uncertain", "group_id": "wf12:archive_loan:test:0086", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0759e75a1c.\\nB means insurance >= 53.\\nC means authorization >= 59.\\nD means restriction >= 48.\\nA means condition >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 32, \\\"insurance\\\": 20, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.19999999999999996, \"true\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0087:choice", "group_id": "wf12:archive_loan:test:0087", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5aca87a2dc.\\nA means condition >= 71.\\nD means restriction >= 86.\\nB means insurance >= 46.\\nC means authorization >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 38, \\\"insurance\\\": 76, \\\"authorization\\\": 64, \\\"restriction\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8d8457\": \"Outcome PROCEED.\", \"candidate_bbcdd8\": \"Outcome REVIEW.\", \"candidate_c092db\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8d8457\": 1.0, \"candidate_bbcdd8\": 0.0, \"candidate_c092db\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0087:score", "group_id": "wf12:archive_loan:test:0087", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5aca87a2dc.\\nA means condition >= 71.\\nD means restriction >= 86.\\nB means insurance >= 46.\\nC means authorization >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 38, \\\"insurance\\\": 76, \\\"authorization\\\": 64, \\\"restriction\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0087:noul", "group_id": "wf12:archive_loan:test:0087", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5aca87a2dc.\\nA means condition >= 71.\\nD means restriction >= 86.\\nB means insurance >= 46.\\nC means authorization >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 38, \\\"insurance\\\": 76, \\\"authorization\\\": 64, \\\"restriction\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0087:uncertain", "group_id": "wf12:archive_loan:test:0087", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5aca87a2dc.\\nA means condition >= 71.\\nD means restriction >= 86.\\nB means insurance >= 46.\\nC means authorization >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 76, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 88}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8d8457\": \"Outcome PROCEED.\", \"candidate_bbcdd8\": \"Outcome REVIEW.\", \"candidate_c092db\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8d8457\": 0.2, \"candidate_bbcdd8\": 0.0, \"candidate_c092db\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0088:choice", "group_id": "wf12:archive_loan:test:0088", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2b97508dc6.\\nC means authorization >= 13.\\nA means condition >= 53.\\nB means insurance >= 90.\\nD means restriction >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": 53, \\\"authorization\\\": 3, \\\"restriction\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_978ce8\": \"Outcome PROCEED.\", \"candidate_6f2d7e\": \"Outcome REVIEW.\", \"candidate_92a874\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_978ce8\": 0.0, \"candidate_6f2d7e\": 1.0, \"candidate_92a874\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0088:score", "group_id": "wf12:archive_loan:test:0088", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2b97508dc6.\\nC means authorization >= 13.\\nA means condition >= 53.\\nB means insurance >= 90.\\nD means restriction >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": 53, \\\"authorization\\\": 3, \\\"restriction\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0088:noul", "group_id": "wf12:archive_loan:test:0088", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2b97508dc6.\\nC means authorization >= 13.\\nA means condition >= 53.\\nB means insurance >= 90.\\nD means restriction >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": 53, \\\"authorization\\\": 3, \\\"restriction\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0088:uncertain", "group_id": "wf12:archive_loan:test:0088", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 2b97508dc6.\\nC means authorization >= 13.\\nA means condition >= 53.\\nB means insurance >= 90.\\nD means restriction >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 9, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 3, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_978ce8\": \"Outcome PROCEED.\", \"candidate_6f2d7e\": \"Outcome REVIEW.\", \"candidate_92a874\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_978ce8\": 0.0, \"candidate_6f2d7e\": 0.09999999999999998, \"candidate_92a874\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0089:choice", "group_id": "wf12:archive_loan:test:0089", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e9f38b579c.\\nA means condition >= 58.\\nB means insurance >= 21.\\nC means authorization >= 31.\\nD means restriction >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 34, \\\"insurance\\\": 27, \\\"authorization\\\": 0, \\\"restriction\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_10109a\": \"Outcome PROCEED.\", \"candidate_ae640e\": \"Outcome REVIEW.\", \"candidate_f899b0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_10109a\": 0.0, \"candidate_ae640e\": 0.0, \"candidate_f899b0\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0089:score", "group_id": "wf12:archive_loan:test:0089", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e9f38b579c.\\nA means condition >= 58.\\nB means insurance >= 21.\\nC means authorization >= 31.\\nD means restriction >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 34, \\\"insurance\\\": 27, \\\"authorization\\\": 0, \\\"restriction\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0089:noul", "group_id": "wf12:archive_loan:test:0089", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e9f38b579c.\\nA means condition >= 58.\\nB means insurance >= 21.\\nC means authorization >= 31.\\nD means restriction >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 34, \\\"insurance\\\": 27, \\\"authorization\\\": 0, \\\"restriction\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0089:uncertain", "group_id": "wf12:archive_loan:test:0089", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e9f38b579c.\\nA means condition >= 58.\\nB means insurance >= 21.\\nC means authorization >= 31.\\nD means restriction >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 34, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 99}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_10109a\": \"Outcome PROCEED.\", \"candidate_ae640e\": \"Outcome REVIEW.\", \"candidate_f899b0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_10109a\": 0.04000000000000001, \"candidate_ae640e\": 0.0, \"candidate_f899b0\": 0.9600000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0090:choice", "group_id": "wf12:archive_loan:test:0090", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 490f5c5e40.\\nC means authorization >= 26.\\nA means condition >= 25.\\nB means insurance >= 85.\\nD means restriction >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 23, \\\"insurance\\\": 100, \\\"authorization\\\": 88, \\\"restriction\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_039828\": \"Outcome PROCEED.\", \"candidate_414a25\": \"Outcome REVIEW.\", \"candidate_df9fc5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_039828\": 1.0, \"candidate_414a25\": 0.0, \"candidate_df9fc5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0090:score", "group_id": "wf12:archive_loan:test:0090", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 490f5c5e40.\\nC means authorization >= 26.\\nA means condition >= 25.\\nB means insurance >= 85.\\nD means restriction >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 23, \\\"insurance\\\": 100, \\\"authorization\\\": 88, \\\"restriction\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0090:noul", "group_id": "wf12:archive_loan:test:0090", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 490f5c5e40.\\nC means authorization >= 26.\\nA means condition >= 25.\\nB means insurance >= 85.\\nD means restriction >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 23, \\\"insurance\\\": 100, \\\"authorization\\\": 88, \\\"restriction\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0090:uncertain", "group_id": "wf12:archive_loan:test:0090", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 490f5c5e40.\\nC means authorization >= 26.\\nA means condition >= 25.\\nB means insurance >= 85.\\nD means restriction >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 23, \\\"insurance\\\": 100, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8, \"1\": 0.12999999999999998, \"2\": 0.06999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0091:choice", "group_id": "wf12:archive_loan:test:0091", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b0eac3dc98.\\nC means authorization >= 39.\\nA means condition >= 41.\\nD means restriction >= 33.\\nB means insurance >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 31, \\\"insurance\\\": 8, \\\"authorization\\\": 5, \\\"restriction\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_628513\": \"Outcome PROCEED.\", \"candidate_7160ff\": \"Outcome REVIEW.\", \"candidate_d8bc5d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_628513\": 0.0, \"candidate_7160ff\": 1.0, \"candidate_d8bc5d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0091:score", "group_id": "wf12:archive_loan:test:0091", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b0eac3dc98.\\nC means authorization >= 39.\\nA means condition >= 41.\\nD means restriction >= 33.\\nB means insurance >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 31, \\\"insurance\\\": 8, \\\"authorization\\\": 5, \\\"restriction\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0091:noul", "group_id": "wf12:archive_loan:test:0091", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b0eac3dc98.\\nC means authorization >= 39.\\nA means condition >= 41.\\nD means restriction >= 33.\\nB means insurance >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 31, \\\"insurance\\\": 8, \\\"authorization\\\": 5, \\\"restriction\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0091:uncertain", "group_id": "wf12:archive_loan:test:0091", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b0eac3dc98.\\nC means authorization >= 39.\\nA means condition >= 41.\\nD means restriction >= 33.\\nB means insurance >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"B\\\": 0.9, \\\"C\\\": 0.95, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.93195, \"1\": 0.053550000000000035, \"2\": 0.014500000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0092:choice", "group_id": "wf12:archive_loan:test:0092", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f41d59b503.\\nA means condition >= 23.\\nC means authorization >= 80.\\nB means insurance >= 26.\\nD means restriction >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 10, \\\"insurance\\\": 66, \\\"authorization\\\": 28, \\\"restriction\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d03e27\": \"Outcome PROCEED.\", \"candidate_265b3d\": \"Outcome REVIEW.\", \"candidate_c3f394\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d03e27\": 0.0, \"candidate_265b3d\": 0.0, \"candidate_c3f394\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0092:score", "group_id": "wf12:archive_loan:test:0092", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f41d59b503.\\nA means condition >= 23.\\nC means authorization >= 80.\\nB means insurance >= 26.\\nD means restriction >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 10, \\\"insurance\\\": 66, \\\"authorization\\\": 28, \\\"restriction\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0092:noul", "group_id": "wf12:archive_loan:test:0092", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f41d59b503.\\nA means condition >= 23.\\nC means authorization >= 80.\\nB means insurance >= 26.\\nD means restriction >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 10, \\\"insurance\\\": 66, \\\"authorization\\\": 28, \\\"restriction\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0092:uncertain", "group_id": "wf12:archive_loan:test:0092", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f41d59b503.\\nA means condition >= 23.\\nC means authorization >= 80.\\nB means insurance >= 26.\\nD means restriction >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 10, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 28, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.65, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0093:choice", "group_id": "wf12:archive_loan:test:0093", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d9d323fa2e.\\nA means condition >= 79.\\nC means authorization >= 77.\\nB means insurance >= 68.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 68, \\\"authorization\\\": 90, \\\"restriction\\\": 58}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6f7539\": \"Outcome PROCEED.\", \"candidate_434bca\": \"Outcome REVIEW.\", \"candidate_8f1ef1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6f7539\": 1.0, \"candidate_434bca\": 0.0, \"candidate_8f1ef1\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0093:score", "group_id": "wf12:archive_loan:test:0093", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d9d323fa2e.\\nA means condition >= 79.\\nC means authorization >= 77.\\nB means insurance >= 68.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 68, \\\"authorization\\\": 90, \\\"restriction\\\": 58}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0093:noul", "group_id": "wf12:archive_loan:test:0093", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d9d323fa2e.\\nA means condition >= 79.\\nC means authorization >= 77.\\nB means insurance >= 68.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 100, \\\"insurance\\\": 68, \\\"authorization\\\": 90, \\\"restriction\\\": 58}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0093:uncertain", "group_id": "wf12:archive_loan:test:0093", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d9d323fa2e.\\nA means condition >= 79.\\nC means authorization >= 77.\\nB means insurance >= 68.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.35, \\\"C\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8455, \"true\": 0.1545}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0094:choice", "group_id": "wf12:archive_loan:test:0094", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 770de2158a.\\nB means insurance >= 65.\\nA means condition >= 79.\\nC means authorization >= 26.\\nD means restriction >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 55, \\\"insurance\\\": 16, \\\"authorization\\\": 13, \\\"restriction\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_800d02\": \"Outcome PROCEED.\", \"candidate_50f4c5\": \"Outcome REVIEW.\", \"candidate_188eed\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_800d02\": 0.0, \"candidate_50f4c5\": 1.0, \"candidate_188eed\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0094:score", "group_id": "wf12:archive_loan:test:0094", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 770de2158a.\\nB means insurance >= 65.\\nA means condition >= 79.\\nC means authorization >= 26.\\nD means restriction >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 55, \\\"insurance\\\": 16, \\\"authorization\\\": 13, \\\"restriction\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0094:noul", "group_id": "wf12:archive_loan:test:0094", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 770de2158a.\\nB means insurance >= 65.\\nA means condition >= 79.\\nC means authorization >= 26.\\nD means restriction >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 55, \\\"insurance\\\": 16, \\\"authorization\\\": 13, \\\"restriction\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0094:uncertain", "group_id": "wf12:archive_loan:test:0094", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 770de2158a.\\nB means insurance >= 65.\\nA means condition >= 79.\\nC means authorization >= 26.\\nD means restriction >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 55, \\\"insurance\\\": 16, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.95, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.050000000000000044, \"true\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0095:choice", "group_id": "wf12:archive_loan:test:0095", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d4b575810a.\\nC means authorization >= 63.\\nB means insurance >= 13.\\nA means condition >= 88.\\nD means restriction >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 10, \\\"authorization\\\": 94, \\\"restriction\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_433a59\": \"Outcome PROCEED.\", \"candidate_bc7fd5\": \"Outcome REVIEW.\", \"candidate_59bcec\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_433a59\": 0.0, \"candidate_bc7fd5\": 0.0, \"candidate_59bcec\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0095:score", "group_id": "wf12:archive_loan:test:0095", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d4b575810a.\\nC means authorization >= 63.\\nB means insurance >= 13.\\nA means condition >= 88.\\nD means restriction >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 10, \\\"authorization\\\": 94, \\\"restriction\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0095:noul", "group_id": "wf12:archive_loan:test:0095", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d4b575810a.\\nC means authorization >= 63.\\nB means insurance >= 13.\\nA means condition >= 88.\\nD means restriction >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 10, \\\"authorization\\\": 94, \\\"restriction\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0095:uncertain", "group_id": "wf12:archive_loan:test:0095", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d4b575810a.\\nC means authorization >= 63.\\nB means insurance >= 13.\\nA means condition >= 88.\\nD means restriction >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 10, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.65, \"true\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0096:choice", "group_id": "wf12:archive_loan:test:0096", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 698e487b04.\\nD means restriction >= 78.\\nC means authorization >= 72.\\nA means condition >= 40.\\nB means insurance >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 24, \\\"authorization\\\": 85, \\\"restriction\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e42bd8\": \"Outcome PROCEED.\", \"candidate_a70022\": \"Outcome REVIEW.\", \"candidate_66a2d9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e42bd8\": 1.0, \"candidate_a70022\": 0.0, \"candidate_66a2d9\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0096:score", "group_id": "wf12:archive_loan:test:0096", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 698e487b04.\\nD means restriction >= 78.\\nC means authorization >= 72.\\nA means condition >= 40.\\nB means insurance >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 24, \\\"authorization\\\": 85, \\\"restriction\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0096:noul", "group_id": "wf12:archive_loan:test:0096", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 698e487b04.\\nD means restriction >= 78.\\nC means authorization >= 72.\\nA means condition >= 40.\\nB means insurance >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 24, \\\"authorization\\\": 85, \\\"restriction\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0096:uncertain", "group_id": "wf12:archive_loan:test:0096", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 698e487b04.\\nD means restriction >= 78.\\nC means authorization >= 72.\\nA means condition >= 40.\\nB means insurance >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 90}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e42bd8\": \"Outcome PROCEED.\", \"candidate_a70022\": \"Outcome REVIEW.\", \"candidate_66a2d9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e42bd8\": 0.12249999999999998, \"candidate_a70022\": 0.0, \"candidate_66a2d9\": 0.8775}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0097:choice", "group_id": "wf12:archive_loan:test:0097", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 85d4c6a3ac.\\nA means condition >= 35.\\nD means restriction >= 74.\\nB means insurance >= 67.\\nC means authorization >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 28, \\\"authorization\\\": 19, \\\"restriction\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0335a4\": \"Outcome PROCEED.\", \"candidate_9c1ff0\": \"Outcome REVIEW.\", \"candidate_9a7bd5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0335a4\": 0.0, \"candidate_9c1ff0\": 1.0, \"candidate_9a7bd5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0097:score", "group_id": "wf12:archive_loan:test:0097", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 85d4c6a3ac.\\nA means condition >= 35.\\nD means restriction >= 74.\\nB means insurance >= 67.\\nC means authorization >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 28, \\\"authorization\\\": 19, \\\"restriction\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0097:noul", "group_id": "wf12:archive_loan:test:0097", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 85d4c6a3ac.\\nA means condition >= 35.\\nD means restriction >= 74.\\nB means insurance >= 67.\\nC means authorization >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 28, \\\"authorization\\\": 19, \\\"restriction\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0097:uncertain", "group_id": "wf12:archive_loan:test:0097", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 85d4c6a3ac.\\nA means condition >= 35.\\nD means restriction >= 74.\\nB means insurance >= 67.\\nC means authorization >= 34.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 28, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.1, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0335a4\": \"Outcome PROCEED.\", \"candidate_9c1ff0\": \"Outcome REVIEW.\", \"candidate_9a7bd5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0335a4\": 0.0, \"candidate_9c1ff0\": 0.9, \"candidate_9a7bd5\": 0.1}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0098:choice", "group_id": "wf12:archive_loan:test:0098", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4a8779db16.\\nD means restriction >= 59.\\nA means condition >= 67.\\nB means insurance >= 17.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 12, \\\"authorization\\\": 77, \\\"restriction\\\": 80}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d3a27f\": \"Outcome PROCEED.\", \"candidate_b92e59\": \"Outcome REVIEW.\", \"candidate_cd90b8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d3a27f\": 0.0, \"candidate_b92e59\": 0.0, \"candidate_cd90b8\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0098:score", "group_id": "wf12:archive_loan:test:0098", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4a8779db16.\\nD means restriction >= 59.\\nA means condition >= 67.\\nB means insurance >= 17.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 12, \\\"authorization\\\": 77, \\\"restriction\\\": 80}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0098:noul", "group_id": "wf12:archive_loan:test:0098", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4a8779db16.\\nD means restriction >= 59.\\nA means condition >= 67.\\nB means insurance >= 17.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 12, \\\"authorization\\\": 77, \\\"restriction\\\": 80}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0098:uncertain", "group_id": "wf12:archive_loan:test:0098", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4a8779db16.\\nD means restriction >= 59.\\nA means condition >= 67.\\nB means insurance >= 17.\\nC means authorization >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 12, \\\"authorization\\\": 77, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d3a27f\": \"Outcome PROCEED.\", \"candidate_b92e59\": \"Outcome REVIEW.\", \"candidate_cd90b8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d3a27f\": 0.27999999999999997, \"candidate_b92e59\": 0.52, \"candidate_cd90b8\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0099:choice", "group_id": "wf12:archive_loan:test:0099", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c08310ebb7.\\nB means insurance >= 76.\\nD means restriction >= 12.\\nC means authorization >= 10.\\nA means condition >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 76, \\\"authorization\\\": 11, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_15551b\": \"Outcome PROCEED.\", \"candidate_da4b9c\": \"Outcome REVIEW.\", \"candidate_37e3a4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_15551b\": 1.0, \"candidate_da4b9c\": 0.0, \"candidate_37e3a4\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0099:score", "group_id": "wf12:archive_loan:test:0099", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c08310ebb7.\\nB means insurance >= 76.\\nD means restriction >= 12.\\nC means authorization >= 10.\\nA means condition >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 76, \\\"authorization\\\": 11, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0099:noul", "group_id": "wf12:archive_loan:test:0099", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c08310ebb7.\\nB means insurance >= 76.\\nD means restriction >= 12.\\nC means authorization >= 10.\\nA means condition >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 76, \\\"authorization\\\": 11, \\\"restriction\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0099:uncertain", "group_id": "wf12:archive_loan:test:0099", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c08310ebb7.\\nB means insurance >= 76.\\nD means restriction >= 12.\\nC means authorization >= 10.\\nA means condition >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 76, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8960000000000001, \"1\": 0.020799999999999968, \"2\": 0.0831999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0100:choice", "group_id": "wf12:archive_loan:test:0100", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3dddafe95c.\\nD means restriction >= 67.\\nC means authorization >= 21.\\nB means insurance >= 88.\\nA means condition >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 69, \\\"authorization\\\": 1, \\\"restriction\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b63251\": \"Outcome PROCEED.\", \"candidate_4326ec\": \"Outcome REVIEW.\", \"candidate_513587\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b63251\": 0.0, \"candidate_4326ec\": 1.0, \"candidate_513587\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0100:score", "group_id": "wf12:archive_loan:test:0100", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3dddafe95c.\\nD means restriction >= 67.\\nC means authorization >= 21.\\nB means insurance >= 88.\\nA means condition >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 69, \\\"authorization\\\": 1, \\\"restriction\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0100:noul", "group_id": "wf12:archive_loan:test:0100", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3dddafe95c.\\nD means restriction >= 67.\\nC means authorization >= 21.\\nB means insurance >= 88.\\nA means condition >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 69, \\\"authorization\\\": 1, \\\"restriction\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0100:uncertain", "group_id": "wf12:archive_loan:test:0100", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3dddafe95c.\\nD means restriction >= 67.\\nC means authorization >= 21.\\nB means insurance >= 88.\\nA means condition >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 69, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.9, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.050000000000000044, \"2\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0101:choice", "group_id": "wf12:archive_loan:test:0101", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 23119373e0.\\nC means authorization >= 38.\\nD means restriction >= 71.\\nA means condition >= 28.\\nB means insurance >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 75, \\\"insurance\\\": 57, \\\"authorization\\\": 76, \\\"restriction\\\": 78}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9cefba\": \"Outcome PROCEED.\", \"candidate_63249a\": \"Outcome REVIEW.\", \"candidate_bb3b68\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9cefba\": 0.0, \"candidate_63249a\": 0.0, \"candidate_bb3b68\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0101:score", "group_id": "wf12:archive_loan:test:0101", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 23119373e0.\\nC means authorization >= 38.\\nD means restriction >= 71.\\nA means condition >= 28.\\nB means insurance >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 75, \\\"insurance\\\": 57, \\\"authorization\\\": 76, \\\"restriction\\\": 78}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0101:noul", "group_id": "wf12:archive_loan:test:0101", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 23119373e0.\\nC means authorization >= 38.\\nD means restriction >= 71.\\nA means condition >= 28.\\nB means insurance >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 75, \\\"insurance\\\": 57, \\\"authorization\\\": 76, \\\"restriction\\\": 78}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0101:uncertain", "group_id": "wf12:archive_loan:test:0101", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 23119373e0.\\nC means authorization >= 38.\\nD means restriction >= 71.\\nA means condition >= 28.\\nB means insurance >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 75, \\\"insurance\\\": 57, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.42250000000000004, \"1\": 0.22749999999999998, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0102:choice", "group_id": "wf12:archive_loan:test:0102", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b67ee12f0f.\\nB means insurance >= 50.\\nD means restriction >= 40.\\nA means condition >= 72.\\nC means authorization >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 89, \\\"authorization\\\": 96, \\\"restriction\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0e0236\": \"Outcome PROCEED.\", \"candidate_456976\": \"Outcome REVIEW.\", \"candidate_43e41e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0e0236\": 1.0, \"candidate_456976\": 0.0, \"candidate_43e41e\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0102:score", "group_id": "wf12:archive_loan:test:0102", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b67ee12f0f.\\nB means insurance >= 50.\\nD means restriction >= 40.\\nA means condition >= 72.\\nC means authorization >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 89, \\\"authorization\\\": 96, \\\"restriction\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0102:noul", "group_id": "wf12:archive_loan:test:0102", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b67ee12f0f.\\nB means insurance >= 50.\\nD means restriction >= 40.\\nA means condition >= 72.\\nC means authorization >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 89, \\\"authorization\\\": 96, \\\"restriction\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0102:uncertain", "group_id": "wf12:archive_loan:test:0102", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b67ee12f0f.\\nB means insurance >= 50.\\nD means restriction >= 40.\\nA means condition >= 72.\\nC means authorization >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 21, \\\"insurance\\\": 89, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.699041095890411, \"true\": 0.3009589041095891}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0103:choice", "group_id": "wf12:archive_loan:test:0103", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 88d4f857f4.\\nA means condition >= 25.\\nB means insurance >= 84.\\nD means restriction >= 62.\\nC means authorization >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 14, \\\"insurance\\\": 84, \\\"authorization\\\": 9, \\\"restriction\\\": 19}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_504a18\": \"Outcome PROCEED.\", \"candidate_9c1d4f\": \"Outcome REVIEW.\", \"candidate_65ec28\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_504a18\": 0.0, \"candidate_9c1d4f\": 1.0, \"candidate_65ec28\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0103:score", "group_id": "wf12:archive_loan:test:0103", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 88d4f857f4.\\nA means condition >= 25.\\nB means insurance >= 84.\\nD means restriction >= 62.\\nC means authorization >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 14, \\\"insurance\\\": 84, \\\"authorization\\\": 9, \\\"restriction\\\": 19}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0103:noul", "group_id": "wf12:archive_loan:test:0103", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 88d4f857f4.\\nA means condition >= 25.\\nB means insurance >= 84.\\nD means restriction >= 62.\\nC means authorization >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 14, \\\"insurance\\\": 84, \\\"authorization\\\": 9, \\\"restriction\\\": 19}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0103:uncertain", "group_id": "wf12:archive_loan:test:0103", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 88d4f857f4.\\nA means condition >= 25.\\nB means insurance >= 84.\\nD means restriction >= 62.\\nC means authorization >= 65.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.1, \\\"C\\\": 0.1, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.059499999999999886, \"true\": 0.9405000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0104:choice", "group_id": "wf12:archive_loan:test:0104", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 947fbdea28.\\nA means condition >= 18.\\nD means restriction >= 75.\\nC means authorization >= 52.\\nB means insurance >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 9, \\\"authorization\\\": 8, \\\"restriction\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_51ee94\": \"Outcome PROCEED.\", \"candidate_a2b172\": \"Outcome REVIEW.\", \"candidate_a47a54\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_51ee94\": 0.0, \"candidate_a2b172\": 0.0, \"candidate_a47a54\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0104:score", "group_id": "wf12:archive_loan:test:0104", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 947fbdea28.\\nA means condition >= 18.\\nD means restriction >= 75.\\nC means authorization >= 52.\\nB means insurance >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 9, \\\"authorization\\\": 8, \\\"restriction\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0104:noul", "group_id": "wf12:archive_loan:test:0104", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 947fbdea28.\\nA means condition >= 18.\\nD means restriction >= 75.\\nC means authorization >= 52.\\nB means insurance >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 42, \\\"insurance\\\": 9, \\\"authorization\\\": 8, \\\"restriction\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0104:uncertain", "group_id": "wf12:archive_loan:test:0104", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 947fbdea28.\\nA means condition >= 18.\\nD means restriction >= 75.\\nC means authorization >= 52.\\nB means insurance >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 9, \\\"authorization\\\": 8, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.19999999999999984, \"true\": 0.8000000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0105:choice", "group_id": "wf12:archive_loan:test:0105", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference db4b199114.\\nC means authorization >= 60.\\nD means restriction >= 38.\\nA means condition >= 21.\\nB means insurance >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 45, \\\"insurance\\\": 90, \\\"authorization\\\": 71, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1b98d8\": \"Outcome PROCEED.\", \"candidate_22892f\": \"Outcome REVIEW.\", \"candidate_e9248c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1b98d8\": 1.0, \"candidate_22892f\": 0.0, \"candidate_e9248c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0105:score", "group_id": "wf12:archive_loan:test:0105", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference db4b199114.\\nC means authorization >= 60.\\nD means restriction >= 38.\\nA means condition >= 21.\\nB means insurance >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 45, \\\"insurance\\\": 90, \\\"authorization\\\": 71, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0105:noul", "group_id": "wf12:archive_loan:test:0105", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference db4b199114.\\nC means authorization >= 60.\\nD means restriction >= 38.\\nA means condition >= 21.\\nB means insurance >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 45, \\\"insurance\\\": 90, \\\"authorization\\\": 71, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0105:uncertain", "group_id": "wf12:archive_loan:test:0105", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference db4b199114.\\nC means authorization >= 60.\\nD means restriction >= 38.\\nA means condition >= 21.\\nB means insurance >= 90.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 45, \\\"insurance\\\": 90, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1b98d8\": \"Outcome PROCEED.\", \"candidate_22892f\": \"Outcome REVIEW.\", \"candidate_e9248c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1b98d8\": 0.8, \"candidate_22892f\": 0.03999999999999998, \"candidate_e9248c\": 0.15999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0106:choice", "group_id": "wf12:archive_loan:test:0106", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b81222ec4b.\\nB means insurance >= 80.\\nC means authorization >= 76.\\nA means condition >= 16.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 82, \\\"authorization\\\": 35, \\\"restriction\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8bbe44\": \"Outcome PROCEED.\", \"candidate_acc020\": \"Outcome REVIEW.\", \"candidate_c7849c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8bbe44\": 0.0, \"candidate_acc020\": 1.0, \"candidate_c7849c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0106:score", "group_id": "wf12:archive_loan:test:0106", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b81222ec4b.\\nB means insurance >= 80.\\nC means authorization >= 76.\\nA means condition >= 16.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 82, \\\"authorization\\\": 35, \\\"restriction\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0106:noul", "group_id": "wf12:archive_loan:test:0106", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b81222ec4b.\\nB means insurance >= 80.\\nC means authorization >= 76.\\nA means condition >= 16.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 82, \\\"authorization\\\": 35, \\\"restriction\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0106:uncertain", "group_id": "wf12:archive_loan:test:0106", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b81222ec4b.\\nB means insurance >= 80.\\nC means authorization >= 76.\\nA means condition >= 16.\\nD means restriction >= 72.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 97, \\\"insurance\\\": 82, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.05, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8bbe44\": \"Outcome PROCEED.\", \"candidate_acc020\": \"Outcome REVIEW.\", \"candidate_c7849c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8bbe44\": 0.05, \"candidate_acc020\": 0.9025, \"candidate_c7849c\": 0.0475}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0107:choice", "group_id": "wf12:archive_loan:test:0107", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c1d1b75dd3.\\nD means restriction >= 57.\\nC means authorization >= 63.\\nA means condition >= 40.\\nB means insurance >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 46, \\\"authorization\\\": 23, \\\"restriction\\\": 97}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_af758e\": \"Outcome PROCEED.\", \"candidate_01713e\": \"Outcome REVIEW.\", \"candidate_3cfee9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_af758e\": 0.0, \"candidate_01713e\": 0.0, \"candidate_3cfee9\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0107:score", "group_id": "wf12:archive_loan:test:0107", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c1d1b75dd3.\\nD means restriction >= 57.\\nC means authorization >= 63.\\nA means condition >= 40.\\nB means insurance >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 46, \\\"authorization\\\": 23, \\\"restriction\\\": 97}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0107:noul", "group_id": "wf12:archive_loan:test:0107", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c1d1b75dd3.\\nD means restriction >= 57.\\nC means authorization >= 63.\\nA means condition >= 40.\\nB means insurance >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 7, \\\"insurance\\\": 46, \\\"authorization\\\": 23, \\\"restriction\\\": 97}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0107:uncertain", "group_id": "wf12:archive_loan:test:0107", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c1d1b75dd3.\\nD means restriction >= 57.\\nC means authorization >= 63.\\nA means condition >= 40.\\nB means insurance >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.65, \\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_af758e\": \"Outcome PROCEED.\", \"candidate_01713e\": \"Outcome REVIEW.\", \"candidate_3cfee9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_af758e\": 0.5648, \"candidate_01713e\": 0.051199999999999975, \"candidate_3cfee9\": 0.38399999999999995}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0108:choice", "group_id": "wf12:archive_loan:test:0108", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7a69b74550.\\nC means authorization >= 45.\\nD means restriction >= 19.\\nA means condition >= 41.\\nB means insurance >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 95, \\\"insurance\\\": 93, \\\"authorization\\\": 54, \\\"restriction\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_db77c8\": \"Outcome PROCEED.\", \"candidate_6e7e01\": \"Outcome REVIEW.\", \"candidate_553b62\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_db77c8\": 1.0, \"candidate_6e7e01\": 0.0, \"candidate_553b62\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0108:score", "group_id": "wf12:archive_loan:test:0108", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7a69b74550.\\nC means authorization >= 45.\\nD means restriction >= 19.\\nA means condition >= 41.\\nB means insurance >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 95, \\\"insurance\\\": 93, \\\"authorization\\\": 54, \\\"restriction\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0108:noul", "group_id": "wf12:archive_loan:test:0108", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7a69b74550.\\nC means authorization >= 45.\\nD means restriction >= 19.\\nA means condition >= 41.\\nB means insurance >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 95, \\\"insurance\\\": 93, \\\"authorization\\\": 54, \\\"restriction\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0108:uncertain", "group_id": "wf12:archive_loan:test:0108", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 7a69b74550.\\nC means authorization >= 45.\\nD means restriction >= 19.\\nA means condition >= 41.\\nB means insurance >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.2, \\\"C\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.2756, \"1\": 0.5504, \"2\": 0.17400000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0109:choice", "group_id": "wf12:archive_loan:test:0109", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0d1ccd5fe1.\\nD means restriction >= 16.\\nA means condition >= 15.\\nB means insurance >= 85.\\nC means authorization >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": 47, \\\"authorization\\\": 66, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cef0a9\": \"Outcome PROCEED.\", \"candidate_55c8ce\": \"Outcome REVIEW.\", \"candidate_c2b565\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cef0a9\": 0.0, \"candidate_55c8ce\": 1.0, \"candidate_c2b565\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0109:score", "group_id": "wf12:archive_loan:test:0109", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0d1ccd5fe1.\\nD means restriction >= 16.\\nA means condition >= 15.\\nB means insurance >= 85.\\nC means authorization >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": 47, \\\"authorization\\\": 66, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0109:noul", "group_id": "wf12:archive_loan:test:0109", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0d1ccd5fe1.\\nD means restriction >= 16.\\nA means condition >= 15.\\nB means insurance >= 85.\\nC means authorization >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": 47, \\\"authorization\\\": 66, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0109:uncertain", "group_id": "wf12:archive_loan:test:0109", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 0d1ccd5fe1.\\nD means restriction >= 16.\\nA means condition >= 15.\\nB means insurance >= 85.\\nC means authorization >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 47, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 13}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.045000000000000005, \"1\": 0.955, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0110:choice", "group_id": "wf12:archive_loan:test:0110", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 430b42e5d7.\\nC means authorization >= 70.\\nB means insurance >= 12.\\nD means restriction >= 17.\\nA means condition >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": 10, \\\"authorization\\\": 75, \\\"restriction\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eb431a\": \"Outcome PROCEED.\", \"candidate_09363a\": \"Outcome REVIEW.\", \"candidate_e89409\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eb431a\": 0.0, \"candidate_09363a\": 0.0, \"candidate_e89409\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0110:score", "group_id": "wf12:archive_loan:test:0110", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 430b42e5d7.\\nC means authorization >= 70.\\nB means insurance >= 12.\\nD means restriction >= 17.\\nA means condition >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": 10, \\\"authorization\\\": 75, \\\"restriction\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0110:noul", "group_id": "wf12:archive_loan:test:0110", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 430b42e5d7.\\nC means authorization >= 70.\\nB means insurance >= 12.\\nD means restriction >= 17.\\nA means condition >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": 10, \\\"authorization\\\": 75, \\\"restriction\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0110:uncertain", "group_id": "wf12:archive_loan:test:0110", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 430b42e5d7.\\nC means authorization >= 70.\\nB means insurance >= 12.\\nD means restriction >= 17.\\nA means condition >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 22}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"C\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.27999999999999997, \"1\": 0.0, \"2\": 0.72}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0111:choice", "group_id": "wf12:archive_loan:test:0111", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b58d60255a.\\nC means authorization >= 90.\\nD means restriction >= 23.\\nA means condition >= 27.\\nB means insurance >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 47, \\\"authorization\\\": 91, \\\"restriction\\\": 62}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a261e\": \"Outcome PROCEED.\", \"candidate_b2ad56\": \"Outcome REVIEW.\", \"candidate_75b7af\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a261e\": 1.0, \"candidate_b2ad56\": 0.0, \"candidate_75b7af\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0111:score", "group_id": "wf12:archive_loan:test:0111", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b58d60255a.\\nC means authorization >= 90.\\nD means restriction >= 23.\\nA means condition >= 27.\\nB means insurance >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 47, \\\"authorization\\\": 91, \\\"restriction\\\": 62}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0111:noul", "group_id": "wf12:archive_loan:test:0111", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b58d60255a.\\nC means authorization >= 90.\\nD means restriction >= 23.\\nA means condition >= 27.\\nB means insurance >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 12, \\\"insurance\\\": 47, \\\"authorization\\\": 91, \\\"restriction\\\": 62}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0111:uncertain", "group_id": "wf12:archive_loan:test:0111", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference b58d60255a.\\nC means authorization >= 90.\\nD means restriction >= 23.\\nA means condition >= 27.\\nB means insurance >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 47, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 62}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8, \"true\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0112:choice", "group_id": "wf12:archive_loan:test:0112", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 92aeb7ad82.\\nA means condition >= 90.\\nB means insurance >= 68.\\nC means authorization >= 65.\\nD means restriction >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": 30, \\\"authorization\\\": 30, \\\"restriction\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6fa69e\": \"Outcome PROCEED.\", \"candidate_44f7c2\": \"Outcome REVIEW.\", \"candidate_57a37d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6fa69e\": 0.0, \"candidate_44f7c2\": 1.0, \"candidate_57a37d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0112:score", "group_id": "wf12:archive_loan:test:0112", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 92aeb7ad82.\\nA means condition >= 90.\\nB means insurance >= 68.\\nC means authorization >= 65.\\nD means restriction >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": 30, \\\"authorization\\\": 30, \\\"restriction\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0112:noul", "group_id": "wf12:archive_loan:test:0112", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 92aeb7ad82.\\nA means condition >= 90.\\nB means insurance >= 68.\\nC means authorization >= 65.\\nD means restriction >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 4, \\\"insurance\\\": 30, \\\"authorization\\\": 30, \\\"restriction\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0112:uncertain", "group_id": "wf12:archive_loan:test:0112", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 92aeb7ad82.\\nA means condition >= 90.\\nB means insurance >= 68.\\nC means authorization >= 65.\\nD means restriction >= 50.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 30, \\\"authorization\\\": 30, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.95, \"true\": 0.05}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0113:choice", "group_id": "wf12:archive_loan:test:0113", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 17b87ddefb.\\nD means restriction >= 38.\\nA means condition >= 13.\\nC means authorization >= 47.\\nB means insurance >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 16, \\\"insurance\\\": 22, \\\"authorization\\\": 16, \\\"restriction\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2d737c\": \"Outcome PROCEED.\", \"candidate_a052f3\": \"Outcome REVIEW.\", \"candidate_6e89ff\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2d737c\": 0.0, \"candidate_a052f3\": 0.0, \"candidate_6e89ff\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0113:score", "group_id": "wf12:archive_loan:test:0113", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 17b87ddefb.\\nD means restriction >= 38.\\nA means condition >= 13.\\nC means authorization >= 47.\\nB means insurance >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 16, \\\"insurance\\\": 22, \\\"authorization\\\": 16, \\\"restriction\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0113:noul", "group_id": "wf12:archive_loan:test:0113", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 17b87ddefb.\\nD means restriction >= 38.\\nA means condition >= 13.\\nC means authorization >= 47.\\nB means insurance >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 16, \\\"insurance\\\": 22, \\\"authorization\\\": 16, \\\"restriction\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0113:uncertain", "group_id": "wf12:archive_loan:test:0113", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 17b87ddefb.\\nD means restriction >= 38.\\nA means condition >= 13.\\nC means authorization >= 47.\\nB means insurance >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 16, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 95}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.28, \"true\": 0.72}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0114:choice", "group_id": "wf12:archive_loan:test:0114", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5138692f59.\\nD means restriction >= 60.\\nC means authorization >= 63.\\nA means condition >= 51.\\nB means insurance >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 31, \\\"authorization\\\": 64, \\\"restriction\\\": 36}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cc2ff4\": \"Outcome PROCEED.\", \"candidate_8bcbb8\": \"Outcome REVIEW.\", \"candidate_095ed6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cc2ff4\": 1.0, \"candidate_8bcbb8\": 0.0, \"candidate_095ed6\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0114:score", "group_id": "wf12:archive_loan:test:0114", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5138692f59.\\nD means restriction >= 60.\\nC means authorization >= 63.\\nA means condition >= 51.\\nB means insurance >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 31, \\\"authorization\\\": 64, \\\"restriction\\\": 36}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0114:noul", "group_id": "wf12:archive_loan:test:0114", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5138692f59.\\nD means restriction >= 60.\\nC means authorization >= 63.\\nA means condition >= 51.\\nB means insurance >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 31, \\\"authorization\\\": 64, \\\"restriction\\\": 36}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0114:uncertain", "group_id": "wf12:archive_loan:test:0114", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 5138692f59.\\nD means restriction >= 60.\\nC means authorization >= 63.\\nA means condition >= 51.\\nB means insurance >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 83, \\\"insurance\\\": 31, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cc2ff4\": \"Outcome PROCEED.\", \"candidate_8bcbb8\": \"Outcome REVIEW.\", \"candidate_095ed6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cc2ff4\": 0.7168000000000001, \"candidate_8bcbb8\": 0.0831999999999999, \"candidate_095ed6\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0115:choice", "group_id": "wf12:archive_loan:test:0115", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c245c65ec3.\\nA means condition >= 65.\\nC means authorization >= 36.\\nB means insurance >= 84.\\nD means restriction >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 64, \\\"insurance\\\": 94, \\\"authorization\\\": 32, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_15d363\": \"Outcome PROCEED.\", \"candidate_a1e143\": \"Outcome REVIEW.\", \"candidate_fa76bc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_15d363\": 0.0, \"candidate_a1e143\": 1.0, \"candidate_fa76bc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0115:score", "group_id": "wf12:archive_loan:test:0115", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c245c65ec3.\\nA means condition >= 65.\\nC means authorization >= 36.\\nB means insurance >= 84.\\nD means restriction >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 64, \\\"insurance\\\": 94, \\\"authorization\\\": 32, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0115:noul", "group_id": "wf12:archive_loan:test:0115", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c245c65ec3.\\nA means condition >= 65.\\nC means authorization >= 36.\\nB means insurance >= 84.\\nD means restriction >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 64, \\\"insurance\\\": 94, \\\"authorization\\\": 32, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0115:uncertain", "group_id": "wf12:archive_loan:test:0115", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c245c65ec3.\\nA means condition >= 65.\\nC means authorization >= 36.\\nB means insurance >= 84.\\nD means restriction >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.1, \\\"C\\\": 0.05, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_15d363\": \"Outcome PROCEED.\", \"candidate_a1e143\": \"Outcome REVIEW.\", \"candidate_fa76bc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_15d363\": 0.00545, \"candidate_a1e143\": 0.09904999999999999, \"candidate_fa76bc\": 0.8955}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0116:choice", "group_id": "wf12:archive_loan:test:0116", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 04e6cc6424.\\nB means insurance >= 52.\\nD means restriction >= 72.\\nC means authorization >= 65.\\nA means condition >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 32, \\\"insurance\\\": 35, \\\"authorization\\\": 68, \\\"restriction\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c9a6f8\": \"Outcome PROCEED.\", \"candidate_af3bb4\": \"Outcome REVIEW.\", \"candidate_bb5678\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c9a6f8\": 0.0, \"candidate_af3bb4\": 0.0, \"candidate_bb5678\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0116:score", "group_id": "wf12:archive_loan:test:0116", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 04e6cc6424.\\nB means insurance >= 52.\\nD means restriction >= 72.\\nC means authorization >= 65.\\nA means condition >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 32, \\\"insurance\\\": 35, \\\"authorization\\\": 68, \\\"restriction\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0116:noul", "group_id": "wf12:archive_loan:test:0116", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 04e6cc6424.\\nB means insurance >= 52.\\nD means restriction >= 72.\\nC means authorization >= 65.\\nA means condition >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 32, \\\"insurance\\\": 35, \\\"authorization\\\": 68, \\\"restriction\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0116:uncertain", "group_id": "wf12:archive_loan:test:0116", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 04e6cc6424.\\nB means insurance >= 52.\\nD means restriction >= 72.\\nC means authorization >= 65.\\nA means condition >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 32, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 68, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c9a6f8\": \"Outcome PROCEED.\", \"candidate_af3bb4\": \"Outcome REVIEW.\", \"candidate_bb5678\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c9a6f8\": 0.8000000000000002, \"candidate_af3bb4\": 0.15999999999999998, \"candidate_bb5678\": 0.039999999999999994}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0117:choice", "group_id": "wf12:archive_loan:test:0117", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e656f95da7.\\nA means condition >= 68.\\nC means authorization >= 64.\\nB means insurance >= 62.\\nD means restriction >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 58, \\\"authorization\\\": 81, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_40a590\": \"Outcome PROCEED.\", \"candidate_81e4be\": \"Outcome REVIEW.\", \"candidate_6edf7f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_40a590\": 1.0, \"candidate_81e4be\": 0.0, \"candidate_6edf7f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0117:score", "group_id": "wf12:archive_loan:test:0117", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e656f95da7.\\nA means condition >= 68.\\nC means authorization >= 64.\\nB means insurance >= 62.\\nD means restriction >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 58, \\\"authorization\\\": 81, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0117:noul", "group_id": "wf12:archive_loan:test:0117", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e656f95da7.\\nA means condition >= 68.\\nC means authorization >= 64.\\nB means insurance >= 62.\\nD means restriction >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 58, \\\"authorization\\\": 81, \\\"restriction\\\": 32}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0117:uncertain", "group_id": "wf12:archive_loan:test:0117", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e656f95da7.\\nA means condition >= 68.\\nC means authorization >= 64.\\nB means insurance >= 62.\\nD means restriction >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 93, \\\"insurance\\\": 58, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.15999999999999998, \"1\": 0.03999999999999998, \"2\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0118:choice", "group_id": "wf12:archive_loan:test:0118", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a332c68fe6.\\nB means insurance >= 55.\\nA means condition >= 74.\\nD means restriction >= 65.\\nC means authorization >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": 6, \\\"authorization\\\": 31, \\\"restriction\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f97021\": \"Outcome PROCEED.\", \"candidate_7556c4\": \"Outcome REVIEW.\", \"candidate_d568d0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f97021\": 0.0, \"candidate_7556c4\": 1.0, \"candidate_d568d0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0118:score", "group_id": "wf12:archive_loan:test:0118", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a332c68fe6.\\nB means insurance >= 55.\\nA means condition >= 74.\\nD means restriction >= 65.\\nC means authorization >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": 6, \\\"authorization\\\": 31, \\\"restriction\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0118:noul", "group_id": "wf12:archive_loan:test:0118", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a332c68fe6.\\nB means insurance >= 55.\\nA means condition >= 74.\\nD means restriction >= 65.\\nC means authorization >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 37, \\\"insurance\\\": 6, \\\"authorization\\\": 31, \\\"restriction\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0118:uncertain", "group_id": "wf12:archive_loan:test:0118", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a332c68fe6.\\nB means insurance >= 55.\\nA means condition >= 74.\\nD means restriction >= 65.\\nC means authorization >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.95, \\\"C\\\": 0.05, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0475125, \"1\": 0.09523749999999997, \"2\": 0.85725}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0119:choice", "group_id": "wf12:archive_loan:test:0119", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4876be3e83.\\nC means authorization >= 73.\\nA means condition >= 68.\\nD means restriction >= 49.\\nB means insurance >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 77, \\\"insurance\\\": 1, \\\"authorization\\\": 100, \\\"restriction\\\": 62}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_06a291\": \"Outcome PROCEED.\", \"candidate_8fb8f8\": \"Outcome REVIEW.\", \"candidate_f29ab9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_06a291\": 0.0, \"candidate_8fb8f8\": 0.0, \"candidate_f29ab9\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0119:score", "group_id": "wf12:archive_loan:test:0119", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4876be3e83.\\nC means authorization >= 73.\\nA means condition >= 68.\\nD means restriction >= 49.\\nB means insurance >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 77, \\\"insurance\\\": 1, \\\"authorization\\\": 100, \\\"restriction\\\": 62}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0119:noul", "group_id": "wf12:archive_loan:test:0119", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4876be3e83.\\nC means authorization >= 73.\\nA means condition >= 68.\\nD means restriction >= 49.\\nB means insurance >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 77, \\\"insurance\\\": 1, \\\"authorization\\\": 100, \\\"restriction\\\": 62}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0119:uncertain", "group_id": "wf12:archive_loan:test:0119", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4876be3e83.\\nC means authorization >= 73.\\nA means condition >= 68.\\nD means restriction >= 49.\\nB means insurance >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 1, \\\"authorization\\\": 100, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.22749999999999998, \"1\": 0.42250000000000004, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0120:choice", "group_id": "wf12:archive_loan:test:0120", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 428e8e7c4d.\\nD means restriction >= 64.\\nA means condition >= 17.\\nC means authorization >= 32.\\nB means insurance >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 96, \\\"insurance\\\": 47, \\\"authorization\\\": 66, \\\"restriction\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2653f5\": \"Outcome PROCEED.\", \"candidate_35ec87\": \"Outcome REVIEW.\", \"candidate_389842\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2653f5\": 1.0, \"candidate_35ec87\": 0.0, \"candidate_389842\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0120:score", "group_id": "wf12:archive_loan:test:0120", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 428e8e7c4d.\\nD means restriction >= 64.\\nA means condition >= 17.\\nC means authorization >= 32.\\nB means insurance >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 96, \\\"insurance\\\": 47, \\\"authorization\\\": 66, \\\"restriction\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0120:noul", "group_id": "wf12:archive_loan:test:0120", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 428e8e7c4d.\\nD means restriction >= 64.\\nA means condition >= 17.\\nC means authorization >= 32.\\nB means insurance >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 96, \\\"insurance\\\": 47, \\\"authorization\\\": 66, \\\"restriction\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0120:uncertain", "group_id": "wf12:archive_loan:test:0120", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 428e8e7c4d.\\nD means restriction >= 64.\\nA means condition >= 17.\\nC means authorization >= 32.\\nB means insurance >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.35, \\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.4296249999999999, \"true\": 0.5703750000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0121:choice", "group_id": "wf12:archive_loan:test:0121", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ca0023c22b.\\nA means condition >= 61.\\nC means authorization >= 89.\\nD means restriction >= 12.\\nB means insurance >= 73.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 44, \\\"insurance\\\": 89, \\\"authorization\\\": 44, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_790c31\": \"Outcome PROCEED.\", \"candidate_cfc670\": \"Outcome REVIEW.\", \"candidate_5bf409\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_790c31\": 0.0, \"candidate_cfc670\": 1.0, \"candidate_5bf409\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0121:score", "group_id": "wf12:archive_loan:test:0121", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ca0023c22b.\\nA means condition >= 61.\\nC means authorization >= 89.\\nD means restriction >= 12.\\nB means insurance >= 73.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 44, \\\"insurance\\\": 89, \\\"authorization\\\": 44, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0121:noul", "group_id": "wf12:archive_loan:test:0121", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ca0023c22b.\\nA means condition >= 61.\\nC means authorization >= 89.\\nD means restriction >= 12.\\nB means insurance >= 73.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 44, \\\"insurance\\\": 89, \\\"authorization\\\": 44, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0121:uncertain", "group_id": "wf12:archive_loan:test:0121", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ca0023c22b.\\nA means condition >= 61.\\nC means authorization >= 89.\\nD means restriction >= 12.\\nB means insurance >= 73.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.95, \\\"C\\\": 0.9, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8694999999999999, \"true\": 0.13050000000000003}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0122:choice", "group_id": "wf12:archive_loan:test:0122", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 88d7c07cf1.\\nB means insurance >= 63.\\nD means restriction >= 42.\\nC means authorization >= 20.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 95, \\\"insurance\\\": 55, \\\"authorization\\\": 5, \\\"restriction\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4859d6\": \"Outcome PROCEED.\", \"candidate_04cbd5\": \"Outcome REVIEW.\", \"candidate_1310f7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4859d6\": 0.0, \"candidate_04cbd5\": 0.0, \"candidate_1310f7\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0122:score", "group_id": "wf12:archive_loan:test:0122", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 88d7c07cf1.\\nB means insurance >= 63.\\nD means restriction >= 42.\\nC means authorization >= 20.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 95, \\\"insurance\\\": 55, \\\"authorization\\\": 5, \\\"restriction\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0122:noul", "group_id": "wf12:archive_loan:test:0122", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 88d7c07cf1.\\nB means insurance >= 63.\\nD means restriction >= 42.\\nC means authorization >= 20.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 95, \\\"insurance\\\": 55, \\\"authorization\\\": 5, \\\"restriction\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0122:uncertain", "group_id": "wf12:archive_loan:test:0122", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 88d7c07cf1.\\nB means insurance >= 63.\\nD means restriction >= 42.\\nC means authorization >= 20.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 55, \\\"authorization\\\": 5, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.19999999999999996, \"true\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0123:choice", "group_id": "wf12:archive_loan:test:0123", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ebc792d923.\\nA means condition >= 65.\\nC means authorization >= 44.\\nB means insurance >= 77.\\nD means restriction >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 51, \\\"insurance\\\": 91, \\\"authorization\\\": 44, \\\"restriction\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3ff4e4\": \"Outcome PROCEED.\", \"candidate_8ad1d5\": \"Outcome REVIEW.\", \"candidate_51c264\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3ff4e4\": 1.0, \"candidate_8ad1d5\": 0.0, \"candidate_51c264\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0123:score", "group_id": "wf12:archive_loan:test:0123", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ebc792d923.\\nA means condition >= 65.\\nC means authorization >= 44.\\nB means insurance >= 77.\\nD means restriction >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 51, \\\"insurance\\\": 91, \\\"authorization\\\": 44, \\\"restriction\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0123:noul", "group_id": "wf12:archive_loan:test:0123", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ebc792d923.\\nA means condition >= 65.\\nC means authorization >= 44.\\nB means insurance >= 77.\\nD means restriction >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 51, \\\"insurance\\\": 91, \\\"authorization\\\": 44, \\\"restriction\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0123:uncertain", "group_id": "wf12:archive_loan:test:0123", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ebc792d923.\\nA means condition >= 65.\\nC means authorization >= 44.\\nB means insurance >= 77.\\nD means restriction >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 51, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 91}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3ff4e4\": \"Outcome PROCEED.\", \"candidate_8ad1d5\": \"Outcome REVIEW.\", \"candidate_51c264\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3ff4e4\": 0.12249999999999998, \"candidate_8ad1d5\": 0.0, \"candidate_51c264\": 0.8775}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0124:choice", "group_id": "wf12:archive_loan:test:0124", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8e524bba98.\\nC means authorization >= 15.\\nB means insurance >= 59.\\nA means condition >= 75.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 96, \\\"insurance\\\": 86, \\\"authorization\\\": 3, \\\"restriction\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_501011\": \"Outcome PROCEED.\", \"candidate_c3dace\": \"Outcome REVIEW.\", \"candidate_897d5c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_501011\": 0.0, \"candidate_c3dace\": 1.0, \"candidate_897d5c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0124:score", "group_id": "wf12:archive_loan:test:0124", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8e524bba98.\\nC means authorization >= 15.\\nB means insurance >= 59.\\nA means condition >= 75.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 96, \\\"insurance\\\": 86, \\\"authorization\\\": 3, \\\"restriction\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0124:noul", "group_id": "wf12:archive_loan:test:0124", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8e524bba98.\\nC means authorization >= 15.\\nB means insurance >= 59.\\nA means condition >= 75.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 96, \\\"insurance\\\": 86, \\\"authorization\\\": 3, \\\"restriction\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0124:uncertain", "group_id": "wf12:archive_loan:test:0124", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8e524bba98.\\nC means authorization >= 15.\\nB means insurance >= 59.\\nA means condition >= 75.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 96, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 50}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_501011\": \"Outcome PROCEED.\", \"candidate_c3dace\": \"Outcome REVIEW.\", \"candidate_897d5c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_501011\": 0.95, \"candidate_c3dace\": 0.050000000000000044, \"candidate_897d5c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0125:choice", "group_id": "wf12:archive_loan:test:0125", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8e06a74ac1.\\nB means insurance >= 36.\\nC means authorization >= 11.\\nD means restriction >= 81.\\nA means condition >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 91, \\\"insurance\\\": 9, \\\"authorization\\\": 1, \\\"restriction\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1fc5b2\": \"Outcome PROCEED.\", \"candidate_2978fb\": \"Outcome REVIEW.\", \"candidate_76e0e2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1fc5b2\": 0.0, \"candidate_2978fb\": 0.0, \"candidate_76e0e2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0125:score", "group_id": "wf12:archive_loan:test:0125", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8e06a74ac1.\\nB means insurance >= 36.\\nC means authorization >= 11.\\nD means restriction >= 81.\\nA means condition >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 91, \\\"insurance\\\": 9, \\\"authorization\\\": 1, \\\"restriction\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0125:noul", "group_id": "wf12:archive_loan:test:0125", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8e06a74ac1.\\nB means insurance >= 36.\\nC means authorization >= 11.\\nD means restriction >= 81.\\nA means condition >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 91, \\\"insurance\\\": 9, \\\"authorization\\\": 1, \\\"restriction\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0125:uncertain", "group_id": "wf12:archive_loan:test:0125", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 8e06a74ac1.\\nB means insurance >= 36.\\nC means authorization >= 11.\\nD means restriction >= 81.\\nA means condition >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.8, \\\"C\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1fc5b2\": \"Outcome PROCEED.\", \"candidate_2978fb\": \"Outcome REVIEW.\", \"candidate_76e0e2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1fc5b2\": 0.1769, \"candidate_2978fb\": 0.5290999999999999, \"candidate_76e0e2\": 0.294}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0126:choice", "group_id": "wf12:archive_loan:test:0126", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 76edcbcfea.\\nB means insurance >= 36.\\nD means restriction >= 79.\\nA means condition >= 72.\\nC means authorization >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 29, \\\"insurance\\\": 48, \\\"authorization\\\": 88, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f9f3b9\": \"Outcome PROCEED.\", \"candidate_7eeb85\": \"Outcome REVIEW.\", \"candidate_40a56e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f9f3b9\": 1.0, \"candidate_7eeb85\": 0.0, \"candidate_40a56e\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0126:score", "group_id": "wf12:archive_loan:test:0126", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 76edcbcfea.\\nB means insurance >= 36.\\nD means restriction >= 79.\\nA means condition >= 72.\\nC means authorization >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 29, \\\"insurance\\\": 48, \\\"authorization\\\": 88, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0126:noul", "group_id": "wf12:archive_loan:test:0126", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 76edcbcfea.\\nB means insurance >= 36.\\nD means restriction >= 79.\\nA means condition >= 72.\\nC means authorization >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 29, \\\"insurance\\\": 48, \\\"authorization\\\": 88, \\\"restriction\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0126:uncertain", "group_id": "wf12:archive_loan:test:0126", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 76edcbcfea.\\nB means insurance >= 36.\\nD means restriction >= 79.\\nA means condition >= 72.\\nC means authorization >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 29, \\\"insurance\\\": 48, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.5369863013698629, \"1\": 0.16205479452054797, \"2\": 0.3009589041095891}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0127:choice", "group_id": "wf12:archive_loan:test:0127", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 97bd7b6f17.\\nA means condition >= 66.\\nC means authorization >= 42.\\nB means insurance >= 62.\\nD means restriction >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 77, \\\"authorization\\\": 29, \\\"restriction\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_212b20\": \"Outcome PROCEED.\", \"candidate_00675f\": \"Outcome REVIEW.\", \"candidate_53b735\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_212b20\": 0.0, \"candidate_00675f\": 1.0, \"candidate_53b735\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0127:score", "group_id": "wf12:archive_loan:test:0127", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 97bd7b6f17.\\nA means condition >= 66.\\nC means authorization >= 42.\\nB means insurance >= 62.\\nD means restriction >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 77, \\\"authorization\\\": 29, \\\"restriction\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0127:noul", "group_id": "wf12:archive_loan:test:0127", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 97bd7b6f17.\\nA means condition >= 66.\\nC means authorization >= 42.\\nB means insurance >= 62.\\nD means restriction >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 77, \\\"authorization\\\": 29, \\\"restriction\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0127:uncertain", "group_id": "wf12:archive_loan:test:0127", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 97bd7b6f17.\\nA means condition >= 66.\\nC means authorization >= 42.\\nB means insurance >= 62.\\nD means restriction >= 64.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 77, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.9, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9, \"1\": 0.08999999999999998, \"2\": 0.009999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0128:choice", "group_id": "wf12:archive_loan:test:0128", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e7ef7ff0ab.\\nD means restriction >= 89.\\nB means insurance >= 70.\\nA means condition >= 18.\\nC means authorization >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 31, \\\"insurance\\\": 51, \\\"authorization\\\": 89, \\\"restriction\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c6de04\": \"Outcome PROCEED.\", \"candidate_95b814\": \"Outcome REVIEW.\", \"candidate_573e17\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c6de04\": 0.0, \"candidate_95b814\": 0.0, \"candidate_573e17\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0128:score", "group_id": "wf12:archive_loan:test:0128", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e7ef7ff0ab.\\nD means restriction >= 89.\\nB means insurance >= 70.\\nA means condition >= 18.\\nC means authorization >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 31, \\\"insurance\\\": 51, \\\"authorization\\\": 89, \\\"restriction\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0128:noul", "group_id": "wf12:archive_loan:test:0128", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e7ef7ff0ab.\\nD means restriction >= 89.\\nB means insurance >= 70.\\nA means condition >= 18.\\nC means authorization >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 31, \\\"insurance\\\": 51, \\\"authorization\\\": 89, \\\"restriction\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0128:uncertain", "group_id": "wf12:archive_loan:test:0128", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e7ef7ff0ab.\\nD means restriction >= 89.\\nB means insurance >= 70.\\nA means condition >= 18.\\nC means authorization >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 89, \\\"restriction\\\": 98}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.0, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0129:choice", "group_id": "wf12:archive_loan:test:0129", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 175757e640.\\nB means insurance >= 48.\\nD means restriction >= 10.\\nC means authorization >= 54.\\nA means condition >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 51, \\\"authorization\\\": 92, \\\"restriction\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bcb702\": \"Outcome PROCEED.\", \"candidate_4b9f81\": \"Outcome REVIEW.\", \"candidate_b674cc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bcb702\": 1.0, \"candidate_4b9f81\": 0.0, \"candidate_b674cc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0129:score", "group_id": "wf12:archive_loan:test:0129", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 175757e640.\\nB means insurance >= 48.\\nD means restriction >= 10.\\nC means authorization >= 54.\\nA means condition >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 51, \\\"authorization\\\": 92, \\\"restriction\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0129:noul", "group_id": "wf12:archive_loan:test:0129", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 175757e640.\\nB means insurance >= 48.\\nD means restriction >= 10.\\nC means authorization >= 54.\\nA means condition >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 28, \\\"insurance\\\": 51, \\\"authorization\\\": 92, \\\"restriction\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0129:uncertain", "group_id": "wf12:archive_loan:test:0129", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 175757e640.\\nB means insurance >= 48.\\nD means restriction >= 10.\\nC means authorization >= 54.\\nA means condition >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.35, \\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.748, \"true\": 0.252}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0130:choice", "group_id": "wf12:archive_loan:test:0130", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d0ed8bf186.\\nD means restriction >= 59.\\nB means insurance >= 66.\\nC means authorization >= 20.\\nA means condition >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 8, \\\"insurance\\\": 73, \\\"authorization\\\": 16, \\\"restriction\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_67867a\": \"Outcome PROCEED.\", \"candidate_bedf18\": \"Outcome REVIEW.\", \"candidate_e81d5a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_67867a\": 0.0, \"candidate_bedf18\": 1.0, \"candidate_e81d5a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0130:score", "group_id": "wf12:archive_loan:test:0130", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d0ed8bf186.\\nD means restriction >= 59.\\nB means insurance >= 66.\\nC means authorization >= 20.\\nA means condition >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 8, \\\"insurance\\\": 73, \\\"authorization\\\": 16, \\\"restriction\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0130:noul", "group_id": "wf12:archive_loan:test:0130", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d0ed8bf186.\\nD means restriction >= 59.\\nB means insurance >= 66.\\nC means authorization >= 20.\\nA means condition >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 8, \\\"insurance\\\": 73, \\\"authorization\\\": 16, \\\"restriction\\\": 55}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0130:uncertain", "group_id": "wf12:archive_loan:test:0130", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d0ed8bf186.\\nD means restriction >= 59.\\nB means insurance >= 66.\\nC means authorization >= 20.\\nA means condition >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 8, \\\"insurance\\\": 73, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.9, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.995, \"true\": 0.004999999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0131:choice", "group_id": "wf12:archive_loan:test:0131", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 79a45e2cac.\\nC means authorization >= 19.\\nD means restriction >= 27.\\nB means insurance >= 81.\\nA means condition >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 13, \\\"insurance\\\": 16, \\\"authorization\\\": 17, \\\"restriction\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_88b540\": \"Outcome PROCEED.\", \"candidate_b9b5a7\": \"Outcome REVIEW.\", \"candidate_09e6db\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_88b540\": 0.0, \"candidate_b9b5a7\": 0.0, \"candidate_09e6db\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0131:score", "group_id": "wf12:archive_loan:test:0131", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 79a45e2cac.\\nC means authorization >= 19.\\nD means restriction >= 27.\\nB means insurance >= 81.\\nA means condition >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 13, \\\"insurance\\\": 16, \\\"authorization\\\": 17, \\\"restriction\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0131:noul", "group_id": "wf12:archive_loan:test:0131", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 79a45e2cac.\\nC means authorization >= 19.\\nD means restriction >= 27.\\nB means insurance >= 81.\\nA means condition >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 13, \\\"insurance\\\": 16, \\\"authorization\\\": 17, \\\"restriction\\\": 68}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0131:uncertain", "group_id": "wf12:archive_loan:test:0131", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 79a45e2cac.\\nC means authorization >= 19.\\nD means restriction >= 27.\\nB means insurance >= 81.\\nA means condition >= 11.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.35, \\\"C\\\": 0.2, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8140000000000001, \"true\": 0.186}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0132:choice", "group_id": "wf12:archive_loan:test:0132", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 86a98e02a2.\\nC means authorization >= 81.\\nA means condition >= 86.\\nB means insurance >= 86.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 89, \\\"insurance\\\": 97, \\\"authorization\\\": 96, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8a084a\": \"Outcome PROCEED.\", \"candidate_90d369\": \"Outcome REVIEW.\", \"candidate_c56e12\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8a084a\": 1.0, \"candidate_90d369\": 0.0, \"candidate_c56e12\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0132:score", "group_id": "wf12:archive_loan:test:0132", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 86a98e02a2.\\nC means authorization >= 81.\\nA means condition >= 86.\\nB means insurance >= 86.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 89, \\\"insurance\\\": 97, \\\"authorization\\\": 96, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0132:noul", "group_id": "wf12:archive_loan:test:0132", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 86a98e02a2.\\nC means authorization >= 81.\\nA means condition >= 86.\\nB means insurance >= 86.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 89, \\\"insurance\\\": 97, \\\"authorization\\\": 96, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0132:uncertain", "group_id": "wf12:archive_loan:test:0132", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 86a98e02a2.\\nC means authorization >= 81.\\nA means condition >= 86.\\nB means insurance >= 86.\\nD means restriction >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 96, \\\"restriction\\\": 5}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8a084a\": \"Outcome PROCEED.\", \"candidate_90d369\": \"Outcome REVIEW.\", \"candidate_c56e12\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8a084a\": 0.8400000000000002, \"candidate_90d369\": 0.15999999999999998, \"candidate_c56e12\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0133:choice", "group_id": "wf12:archive_loan:test:0133", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d91e1e11bc.\\nB means insurance >= 60.\\nC means authorization >= 66.\\nA means condition >= 24.\\nD means restriction >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": 45, \\\"authorization\\\": 64, \\\"restriction\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_07c6b5\": \"Outcome PROCEED.\", \"candidate_816d2b\": \"Outcome REVIEW.\", \"candidate_4caa8c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_07c6b5\": 0.0, \"candidate_816d2b\": 1.0, \"candidate_4caa8c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0133:score", "group_id": "wf12:archive_loan:test:0133", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d91e1e11bc.\\nB means insurance >= 60.\\nC means authorization >= 66.\\nA means condition >= 24.\\nD means restriction >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": 45, \\\"authorization\\\": 64, \\\"restriction\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0133:noul", "group_id": "wf12:archive_loan:test:0133", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d91e1e11bc.\\nB means insurance >= 60.\\nC means authorization >= 66.\\nA means condition >= 24.\\nD means restriction >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 11, \\\"insurance\\\": 45, \\\"authorization\\\": 64, \\\"restriction\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0133:uncertain", "group_id": "wf12:archive_loan:test:0133", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference d91e1e11bc.\\nB means insurance >= 60.\\nC means authorization >= 66.\\nA means condition >= 24.\\nD means restriction >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 45, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 37}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_07c6b5\": \"Outcome PROCEED.\", \"candidate_816d2b\": \"Outcome REVIEW.\", \"candidate_4caa8c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_07c6b5\": 0.045000000000000005, \"candidate_816d2b\": 0.955, \"candidate_4caa8c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0134:choice", "group_id": "wf12:archive_loan:test:0134", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 10422a6b50.\\nA means condition >= 49.\\nD means restriction >= 64.\\nB means insurance >= 10.\\nC means authorization >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 43, \\\"authorization\\\": 7, \\\"restriction\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7e9420\": \"Outcome PROCEED.\", \"candidate_2838f4\": \"Outcome REVIEW.\", \"candidate_ee1529\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7e9420\": 0.0, \"candidate_2838f4\": 0.0, \"candidate_ee1529\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0134:score", "group_id": "wf12:archive_loan:test:0134", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 10422a6b50.\\nA means condition >= 49.\\nD means restriction >= 64.\\nB means insurance >= 10.\\nC means authorization >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 43, \\\"authorization\\\": 7, \\\"restriction\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0134:noul", "group_id": "wf12:archive_loan:test:0134", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 10422a6b50.\\nA means condition >= 49.\\nD means restriction >= 64.\\nB means insurance >= 10.\\nC means authorization >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": 43, \\\"authorization\\\": 7, \\\"restriction\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0134:uncertain", "group_id": "wf12:archive_loan:test:0134", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 10422a6b50.\\nA means condition >= 49.\\nD means restriction >= 64.\\nB means insurance >= 10.\\nC means authorization >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 5, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 7, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7e9420\": \"Outcome PROCEED.\", \"candidate_2838f4\": \"Outcome REVIEW.\", \"candidate_ee1529\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7e9420\": 0.0, \"candidate_2838f4\": 0.19999999999999996, \"candidate_ee1529\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0135:choice", "group_id": "wf12:archive_loan:test:0135", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e7999611fa.\\nB means insurance >= 21.\\nC means authorization >= 79.\\nD means restriction >= 37.\\nA means condition >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 64, \\\"insurance\\\": 32, \\\"authorization\\\": 91, \\\"restriction\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1fc406\": \"Outcome PROCEED.\", \"candidate_2ec646\": \"Outcome REVIEW.\", \"candidate_9ff657\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1fc406\": 1.0, \"candidate_2ec646\": 0.0, \"candidate_9ff657\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0135:score", "group_id": "wf12:archive_loan:test:0135", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e7999611fa.\\nB means insurance >= 21.\\nC means authorization >= 79.\\nD means restriction >= 37.\\nA means condition >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 64, \\\"insurance\\\": 32, \\\"authorization\\\": 91, \\\"restriction\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0135:noul", "group_id": "wf12:archive_loan:test:0135", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e7999611fa.\\nB means insurance >= 21.\\nC means authorization >= 79.\\nD means restriction >= 37.\\nA means condition >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 64, \\\"insurance\\\": 32, \\\"authorization\\\": 91, \\\"restriction\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0135:uncertain", "group_id": "wf12:archive_loan:test:0135", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference e7999611fa.\\nB means insurance >= 21.\\nC means authorization >= 79.\\nD means restriction >= 37.\\nA means condition >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 91, \\\"restriction\\\": 27}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8379452054794521, \"1\": 0.16205479452054797, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0136:choice", "group_id": "wf12:archive_loan:test:0136", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 46241f1127.\\nA means condition >= 24.\\nC means authorization >= 72.\\nD means restriction >= 87.\\nB means insurance >= 67.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 86, \\\"authorization\\\": 57, \\\"restriction\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d0f81a\": \"Outcome PROCEED.\", \"candidate_78f91a\": \"Outcome REVIEW.\", \"candidate_b983f7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d0f81a\": 0.0, \"candidate_78f91a\": 1.0, \"candidate_b983f7\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0136:score", "group_id": "wf12:archive_loan:test:0136", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 46241f1127.\\nA means condition >= 24.\\nC means authorization >= 72.\\nD means restriction >= 87.\\nB means insurance >= 67.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 86, \\\"authorization\\\": 57, \\\"restriction\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0136:noul", "group_id": "wf12:archive_loan:test:0136", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 46241f1127.\\nA means condition >= 24.\\nC means authorization >= 72.\\nD means restriction >= 87.\\nB means insurance >= 67.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 86, \\\"authorization\\\": 57, \\\"restriction\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0136:uncertain", "group_id": "wf12:archive_loan:test:0136", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 46241f1127.\\nA means condition >= 24.\\nC means authorization >= 72.\\nD means restriction >= 87.\\nB means insurance >= 67.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 86, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 22}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"C\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9, \"1\": 0.09999999999999998, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0137:choice", "group_id": "wf12:archive_loan:test:0137", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3343797180.\\nA means condition >= 32.\\nC means authorization >= 63.\\nB means insurance >= 64.\\nD means restriction >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 6, \\\"authorization\\\": 0, \\\"restriction\\\": 36}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6f5a3a\": \"Outcome PROCEED.\", \"candidate_c69468\": \"Outcome REVIEW.\", \"candidate_d2341f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6f5a3a\": 0.0, \"candidate_c69468\": 0.0, \"candidate_d2341f\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0137:score", "group_id": "wf12:archive_loan:test:0137", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3343797180.\\nA means condition >= 32.\\nC means authorization >= 63.\\nB means insurance >= 64.\\nD means restriction >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 6, \\\"authorization\\\": 0, \\\"restriction\\\": 36}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0137:noul", "group_id": "wf12:archive_loan:test:0137", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3343797180.\\nA means condition >= 32.\\nC means authorization >= 63.\\nB means insurance >= 64.\\nD means restriction >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 6, \\\"authorization\\\": 0, \\\"restriction\\\": 36}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0137:uncertain", "group_id": "wf12:archive_loan:test:0137", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 3343797180.\\nA means condition >= 32.\\nC means authorization >= 63.\\nB means insurance >= 64.\\nD means restriction >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 1, \\\"insurance\\\": 6, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.19999999999999996, \"2\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0138:choice", "group_id": "wf12:archive_loan:test:0138", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 83b3a49975.\\nD means restriction >= 83.\\nC means authorization >= 34.\\nA means condition >= 37.\\nB means insurance >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 36, \\\"insurance\\\": 77, \\\"authorization\\\": 38, \\\"restriction\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bc4f1e\": \"Outcome PROCEED.\", \"candidate_6fbdf2\": \"Outcome REVIEW.\", \"candidate_35b858\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bc4f1e\": 1.0, \"candidate_6fbdf2\": 0.0, \"candidate_35b858\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0138:score", "group_id": "wf12:archive_loan:test:0138", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 83b3a49975.\\nD means restriction >= 83.\\nC means authorization >= 34.\\nA means condition >= 37.\\nB means insurance >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 36, \\\"insurance\\\": 77, \\\"authorization\\\": 38, \\\"restriction\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0138:noul", "group_id": "wf12:archive_loan:test:0138", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 83b3a49975.\\nD means restriction >= 83.\\nC means authorization >= 34.\\nA means condition >= 37.\\nB means insurance >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 36, \\\"insurance\\\": 77, \\\"authorization\\\": 38, \\\"restriction\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0138:uncertain", "group_id": "wf12:archive_loan:test:0138", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 83b3a49975.\\nD means restriction >= 83.\\nC means authorization >= 34.\\nA means condition >= 37.\\nB means insurance >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.2, \\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.6955, \"true\": 0.3045}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0139:choice", "group_id": "wf12:archive_loan:test:0139", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9d26ef931a.\\nC means authorization >= 14.\\nD means restriction >= 27.\\nA means condition >= 45.\\nB means insurance >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 38, \\\"authorization\\\": 4, \\\"restriction\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d88b7d\": \"Outcome PROCEED.\", \"candidate_3c6081\": \"Outcome REVIEW.\", \"candidate_296112\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d88b7d\": 0.0, \"candidate_3c6081\": 1.0, \"candidate_296112\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0139:score", "group_id": "wf12:archive_loan:test:0139", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9d26ef931a.\\nC means authorization >= 14.\\nD means restriction >= 27.\\nA means condition >= 45.\\nB means insurance >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 38, \\\"authorization\\\": 4, \\\"restriction\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0139:noul", "group_id": "wf12:archive_loan:test:0139", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9d26ef931a.\\nC means authorization >= 14.\\nD means restriction >= 27.\\nA means condition >= 45.\\nB means insurance >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 40, \\\"insurance\\\": 38, \\\"authorization\\\": 4, \\\"restriction\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0139:uncertain", "group_id": "wf12:archive_loan:test:0139", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 9d26ef931a.\\nC means authorization >= 14.\\nD means restriction >= 27.\\nA means condition >= 45.\\nB means insurance >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.1, \\\"C\\\": 0.1, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.059499999999999886, \"true\": 0.9405000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0140:choice", "group_id": "wf12:archive_loan:test:0140", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 59db191721.\\nC means authorization >= 69.\\nD means restriction >= 60.\\nA means condition >= 51.\\nB means insurance >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 2, \\\"authorization\\\": 58, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a85746\": \"Outcome PROCEED.\", \"candidate_3c3d3f\": \"Outcome REVIEW.\", \"candidate_f44266\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a85746\": 0.0, \"candidate_3c3d3f\": 0.0, \"candidate_f44266\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0140:score", "group_id": "wf12:archive_loan:test:0140", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 59db191721.\\nC means authorization >= 69.\\nD means restriction >= 60.\\nA means condition >= 51.\\nB means insurance >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 2, \\\"authorization\\\": 58, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0140:noul", "group_id": "wf12:archive_loan:test:0140", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 59db191721.\\nC means authorization >= 69.\\nD means restriction >= 60.\\nA means condition >= 51.\\nB means insurance >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 2, \\\"authorization\\\": 58, \\\"restriction\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0140:uncertain", "group_id": "wf12:archive_loan:test:0140", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 59db191721.\\nC means authorization >= 69.\\nD means restriction >= 60.\\nA means condition >= 51.\\nB means insurance >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 2, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.19999999999999996, \"true\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0141:choice", "group_id": "wf12:archive_loan:test:0141", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference da5dd22a8d.\\nD means restriction >= 20.\\nB means insurance >= 32.\\nC means authorization >= 54.\\nA means condition >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 57, \\\"insurance\\\": 90, \\\"authorization\\\": 86, \\\"restriction\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_aa4c68\": \"Outcome PROCEED.\", \"candidate_7c4fb9\": \"Outcome REVIEW.\", \"candidate_293b90\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_aa4c68\": 1.0, \"candidate_7c4fb9\": 0.0, \"candidate_293b90\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0141:score", "group_id": "wf12:archive_loan:test:0141", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference da5dd22a8d.\\nD means restriction >= 20.\\nB means insurance >= 32.\\nC means authorization >= 54.\\nA means condition >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 57, \\\"insurance\\\": 90, \\\"authorization\\\": 86, \\\"restriction\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0141:noul", "group_id": "wf12:archive_loan:test:0141", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference da5dd22a8d.\\nD means restriction >= 20.\\nB means insurance >= 32.\\nC means authorization >= 54.\\nA means condition >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 57, \\\"insurance\\\": 90, \\\"authorization\\\": 86, \\\"restriction\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0141:uncertain", "group_id": "wf12:archive_loan:test:0141", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference da5dd22a8d.\\nD means restriction >= 20.\\nB means insurance >= 32.\\nC means authorization >= 54.\\nA means condition >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 57, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 86, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_aa4c68\": \"Outcome PROCEED.\", \"candidate_7c4fb9\": \"Outcome REVIEW.\", \"candidate_293b90\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_aa4c68\": 0.9324000000000001, \"candidate_7c4fb9\": 0.0, \"candidate_293b90\": 0.06759999999999992}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0142:choice", "group_id": "wf12:archive_loan:test:0142", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4157ba53d2.\\nB means insurance >= 41.\\nC means authorization >= 45.\\nA means condition >= 70.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 88, \\\"insurance\\\": 61, \\\"authorization\\\": 7, \\\"restriction\\\": 48}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_31e601\": \"Outcome PROCEED.\", \"candidate_967990\": \"Outcome REVIEW.\", \"candidate_ee97b6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_31e601\": 0.0, \"candidate_967990\": 1.0, \"candidate_ee97b6\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0142:score", "group_id": "wf12:archive_loan:test:0142", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4157ba53d2.\\nB means insurance >= 41.\\nC means authorization >= 45.\\nA means condition >= 70.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 88, \\\"insurance\\\": 61, \\\"authorization\\\": 7, \\\"restriction\\\": 48}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0142:noul", "group_id": "wf12:archive_loan:test:0142", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4157ba53d2.\\nB means insurance >= 41.\\nC means authorization >= 45.\\nA means condition >= 70.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 88, \\\"insurance\\\": 61, \\\"authorization\\\": 7, \\\"restriction\\\": 48}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0142:uncertain", "group_id": "wf12:archive_loan:test:0142", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 4157ba53d2.\\nB means insurance >= 41.\\nC means authorization >= 45.\\nA means condition >= 70.\\nD means restriction >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 61, \\\"authorization\\\": 7, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_31e601\": \"Outcome PROCEED.\", \"candidate_967990\": \"Outcome REVIEW.\", \"candidate_ee97b6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_31e601\": 0.0, \"candidate_967990\": 0.09999999999999998, \"candidate_ee97b6\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0143:choice", "group_id": "wf12:archive_loan:test:0143", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6d7f8414cb.\\nA means condition >= 44.\\nC means authorization >= 47.\\nD means restriction >= 20.\\nB means insurance >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 68, \\\"authorization\\\": 2, \\\"restriction\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_565560\": \"Outcome PROCEED.\", \"candidate_e28e4c\": \"Outcome REVIEW.\", \"candidate_03c9c4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_565560\": 0.0, \"candidate_e28e4c\": 0.0, \"candidate_03c9c4\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0143:score", "group_id": "wf12:archive_loan:test:0143", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6d7f8414cb.\\nA means condition >= 44.\\nC means authorization >= 47.\\nD means restriction >= 20.\\nB means insurance >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 68, \\\"authorization\\\": 2, \\\"restriction\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0143:noul", "group_id": "wf12:archive_loan:test:0143", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6d7f8414cb.\\nA means condition >= 44.\\nC means authorization >= 47.\\nD means restriction >= 20.\\nB means insurance >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 94, \\\"insurance\\\": 68, \\\"authorization\\\": 2, \\\"restriction\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0143:uncertain", "group_id": "wf12:archive_loan:test:0143", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 6d7f8414cb.\\nA means condition >= 44.\\nC means authorization >= 47.\\nD means restriction >= 20.\\nB means insurance >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 68, \\\"authorization\\\": 2, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_565560\": \"Outcome PROCEED.\", \"candidate_e28e4c\": \"Outcome REVIEW.\", \"candidate_03c9c4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_565560\": 0.0, \"candidate_e28e4c\": 0.8, \"candidate_03c9c4\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0144:choice", "group_id": "wf12:archive_loan:test:0144", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f87d444cb2.\\nC means authorization >= 44.\\nB means insurance >= 27.\\nD means restriction >= 44.\\nA means condition >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": 92, \\\"authorization\\\": 95, \\\"restriction\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0e170a\": \"Outcome PROCEED.\", \"candidate_608149\": \"Outcome REVIEW.\", \"candidate_a1d923\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0e170a\": 1.0, \"candidate_608149\": 0.0, \"candidate_a1d923\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0144:score", "group_id": "wf12:archive_loan:test:0144", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f87d444cb2.\\nC means authorization >= 44.\\nB means insurance >= 27.\\nD means restriction >= 44.\\nA means condition >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": 92, \\\"authorization\\\": 95, \\\"restriction\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0144:noul", "group_id": "wf12:archive_loan:test:0144", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f87d444cb2.\\nC means authorization >= 44.\\nB means insurance >= 27.\\nD means restriction >= 44.\\nA means condition >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": 92, \\\"authorization\\\": 95, \\\"restriction\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0144:uncertain", "group_id": "wf12:archive_loan:test:0144", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f87d444cb2.\\nC means authorization >= 44.\\nB means insurance >= 27.\\nD means restriction >= 44.\\nA means condition >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 92, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 4}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.2, \"1\": 0.8, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0145:choice", "group_id": "wf12:archive_loan:test:0145", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f10837b460.\\nB means insurance >= 77.\\nD means restriction >= 44.\\nC means authorization >= 12.\\nA means condition >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 59, \\\"insurance\\\": 60, \\\"authorization\\\": 9, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fe3ffb\": \"Outcome PROCEED.\", \"candidate_7cfe56\": \"Outcome REVIEW.\", \"candidate_038063\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fe3ffb\": 0.0, \"candidate_7cfe56\": 1.0, \"candidate_038063\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0145:score", "group_id": "wf12:archive_loan:test:0145", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f10837b460.\\nB means insurance >= 77.\\nD means restriction >= 44.\\nC means authorization >= 12.\\nA means condition >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 59, \\\"insurance\\\": 60, \\\"authorization\\\": 9, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0145:noul", "group_id": "wf12:archive_loan:test:0145", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f10837b460.\\nB means insurance >= 77.\\nD means restriction >= 44.\\nC means authorization >= 12.\\nA means condition >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 59, \\\"insurance\\\": 60, \\\"authorization\\\": 9, \\\"restriction\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0145:uncertain", "group_id": "wf12:archive_loan:test:0145", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f10837b460.\\nB means insurance >= 77.\\nD means restriction >= 44.\\nC means authorization >= 12.\\nA means condition >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 59, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 8}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.1, \"1\": 0.9, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0146:choice", "group_id": "wf12:archive_loan:test:0146", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 965a608891.\\nA means condition >= 17.\\nB means insurance >= 30.\\nC means authorization >= 40.\\nD means restriction >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 60, \\\"insurance\\\": 11, \\\"authorization\\\": 37, \\\"restriction\\\": 86}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bfaa70\": \"Outcome PROCEED.\", \"candidate_33bcb1\": \"Outcome REVIEW.\", \"candidate_4d0fe1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bfaa70\": 0.0, \"candidate_33bcb1\": 0.0, \"candidate_4d0fe1\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0146:score", "group_id": "wf12:archive_loan:test:0146", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 965a608891.\\nA means condition >= 17.\\nB means insurance >= 30.\\nC means authorization >= 40.\\nD means restriction >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 60, \\\"insurance\\\": 11, \\\"authorization\\\": 37, \\\"restriction\\\": 86}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0146:noul", "group_id": "wf12:archive_loan:test:0146", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 965a608891.\\nA means condition >= 17.\\nB means insurance >= 30.\\nC means authorization >= 40.\\nD means restriction >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 60, \\\"insurance\\\": 11, \\\"authorization\\\": 37, \\\"restriction\\\": 86}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0146:uncertain", "group_id": "wf12:archive_loan:test:0146", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 965a608891.\\nA means condition >= 17.\\nB means insurance >= 30.\\nC means authorization >= 40.\\nD means restriction >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 60, \\\"insurance\\\": 11, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.52, \"1\": 0.12999999999999998, \"2\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0147:choice", "group_id": "wf12:archive_loan:test:0147", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference bf12fc31b5.\\nD means restriction >= 79.\\nC means authorization >= 74.\\nA means condition >= 31.\\nB means insurance >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": 45, \\\"authorization\\\": 97, \\\"restriction\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cbcb41\": \"Outcome PROCEED.\", \"candidate_c0bbaf\": \"Outcome REVIEW.\", \"candidate_d7be43\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cbcb41\": 1.0, \"candidate_c0bbaf\": 0.0, \"candidate_d7be43\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0147:score", "group_id": "wf12:archive_loan:test:0147", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference bf12fc31b5.\\nD means restriction >= 79.\\nC means authorization >= 74.\\nA means condition >= 31.\\nB means insurance >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": 45, \\\"authorization\\\": 97, \\\"restriction\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0147:noul", "group_id": "wf12:archive_loan:test:0147", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference bf12fc31b5.\\nD means restriction >= 79.\\nC means authorization >= 74.\\nA means condition >= 31.\\nB means insurance >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": 45, \\\"authorization\\\": 97, \\\"restriction\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0147:uncertain", "group_id": "wf12:archive_loan:test:0147", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference bf12fc31b5.\\nD means restriction >= 79.\\nC means authorization >= 74.\\nA means condition >= 31.\\nB means insurance >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 97, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.93, \"true\": 0.06999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0148:choice", "group_id": "wf12:archive_loan:test:0148", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c3a7254058.\\nC means authorization >= 77.\\nD means restriction >= 13.\\nB means insurance >= 43.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 82, \\\"authorization\\\": 50, \\\"restriction\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a93450\": \"Outcome PROCEED.\", \"candidate_666d05\": \"Outcome REVIEW.\", \"candidate_ce8dda\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a93450\": 0.0, \"candidate_666d05\": 1.0, \"candidate_ce8dda\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0148:score", "group_id": "wf12:archive_loan:test:0148", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c3a7254058.\\nC means authorization >= 77.\\nD means restriction >= 13.\\nB means insurance >= 43.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 82, \\\"authorization\\\": 50, \\\"restriction\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0148:noul", "group_id": "wf12:archive_loan:test:0148", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c3a7254058.\\nC means authorization >= 77.\\nD means restriction >= 13.\\nB means insurance >= 43.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 82, \\\"authorization\\\": 50, \\\"restriction\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0148:uncertain", "group_id": "wf12:archive_loan:test:0148", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c3a7254058.\\nC means authorization >= 77.\\nD means restriction >= 13.\\nB means insurance >= 43.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.9, \\\"C\\\": 0.05, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09275000000000011, \"true\": 0.9072499999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0149:choice", "group_id": "wf12:archive_loan:test:0149", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f0b4d5fdd9.\\nD means restriction >= 57.\\nA means condition >= 89.\\nB means insurance >= 40.\\nC means authorization >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 57, \\\"authorization\\\": 66, \\\"restriction\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b34737\": \"Outcome PROCEED.\", \"candidate_472f09\": \"Outcome REVIEW.\", \"candidate_defb96\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b34737\": 0.0, \"candidate_472f09\": 0.0, \"candidate_defb96\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0149:score", "group_id": "wf12:archive_loan:test:0149", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f0b4d5fdd9.\\nD means restriction >= 57.\\nA means condition >= 89.\\nB means insurance >= 40.\\nC means authorization >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 57, \\\"authorization\\\": 66, \\\"restriction\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0149:noul", "group_id": "wf12:archive_loan:test:0149", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f0b4d5fdd9.\\nD means restriction >= 57.\\nA means condition >= 89.\\nB means insurance >= 40.\\nC means authorization >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 57, \\\"authorization\\\": 66, \\\"restriction\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0149:uncertain", "group_id": "wf12:archive_loan:test:0149", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference f0b4d5fdd9.\\nD means restriction >= 57.\\nA means condition >= 89.\\nB means insurance >= 40.\\nC means authorization >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.8, \\\"C\\\": 0.2, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.45399999999999996, \"true\": 0.546}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0150:choice", "group_id": "wf12:archive_loan:test:0150", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference aa9e783b3f.\\nC means authorization >= 39.\\nD means restriction >= 83.\\nA means condition >= 72.\\nB means insurance >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 73, \\\"authorization\\\": 55, \\\"restriction\\\": 92}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b04e9e\": \"Outcome PROCEED.\", \"candidate_72fff1\": \"Outcome REVIEW.\", \"candidate_fa7dfc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b04e9e\": 1.0, \"candidate_72fff1\": 0.0, \"candidate_fa7dfc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0150:score", "group_id": "wf12:archive_loan:test:0150", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference aa9e783b3f.\\nC means authorization >= 39.\\nD means restriction >= 83.\\nA means condition >= 72.\\nB means insurance >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 73, \\\"authorization\\\": 55, \\\"restriction\\\": 92}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0150:noul", "group_id": "wf12:archive_loan:test:0150", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference aa9e783b3f.\\nC means authorization >= 39.\\nD means restriction >= 83.\\nA means condition >= 72.\\nB means insurance >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 73, \\\"authorization\\\": 55, \\\"restriction\\\": 92}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0150:uncertain", "group_id": "wf12:archive_loan:test:0150", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference aa9e783b3f.\\nC means authorization >= 39.\\nD means restriction >= 83.\\nA means condition >= 72.\\nB means insurance >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 73, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b04e9e\": \"Outcome PROCEED.\", \"candidate_72fff1\": \"Outcome REVIEW.\", \"candidate_fa7dfc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b04e9e\": 0.8, \"candidate_72fff1\": 0.12999999999999998, \"candidate_fa7dfc\": 0.06999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0151:choice", "group_id": "wf12:archive_loan:test:0151", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cead6b138b.\\nB means insurance >= 42.\\nC means authorization >= 35.\\nD means restriction >= 55.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 26, \\\"insurance\\\": 54, \\\"authorization\\\": 31, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_77c399\": \"Outcome PROCEED.\", \"candidate_beec6f\": \"Outcome REVIEW.\", \"candidate_715baf\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_77c399\": 0.0, \"candidate_beec6f\": 1.0, \"candidate_715baf\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0151:score", "group_id": "wf12:archive_loan:test:0151", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cead6b138b.\\nB means insurance >= 42.\\nC means authorization >= 35.\\nD means restriction >= 55.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 26, \\\"insurance\\\": 54, \\\"authorization\\\": 31, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0151:noul", "group_id": "wf12:archive_loan:test:0151", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cead6b138b.\\nB means insurance >= 42.\\nC means authorization >= 35.\\nD means restriction >= 55.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 26, \\\"insurance\\\": 54, \\\"authorization\\\": 31, \\\"restriction\\\": 5}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0151:uncertain", "group_id": "wf12:archive_loan:test:0151", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference cead6b138b.\\nB means insurance >= 42.\\nC means authorization >= 35.\\nD means restriction >= 55.\\nA means condition >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 26, \\\"insurance\\\": 54, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.05, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_77c399\": \"Outcome PROCEED.\", \"candidate_beec6f\": \"Outcome REVIEW.\", \"candidate_715baf\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_77c399\": 0.05, \"candidate_beec6f\": 0.04750000000000004, \"candidate_715baf\": 0.9025}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0152:choice", "group_id": "wf12:archive_loan:test:0152", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c3403a4922.\\nC means authorization >= 23.\\nA means condition >= 78.\\nD means restriction >= 71.\\nB means insurance >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 87, \\\"insurance\\\": 72, \\\"authorization\\\": 68, \\\"restriction\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c7e006\": \"Outcome PROCEED.\", \"candidate_8404c7\": \"Outcome REVIEW.\", \"candidate_6751ef\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c7e006\": 0.0, \"candidate_8404c7\": 0.0, \"candidate_6751ef\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0152:score", "group_id": "wf12:archive_loan:test:0152", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c3403a4922.\\nC means authorization >= 23.\\nA means condition >= 78.\\nD means restriction >= 71.\\nB means insurance >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 87, \\\"insurance\\\": 72, \\\"authorization\\\": 68, \\\"restriction\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0152:noul", "group_id": "wf12:archive_loan:test:0152", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c3403a4922.\\nC means authorization >= 23.\\nA means condition >= 78.\\nD means restriction >= 71.\\nB means insurance >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 87, \\\"insurance\\\": 72, \\\"authorization\\\": 68, \\\"restriction\\\": 79}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0152:uncertain", "group_id": "wf12:archive_loan:test:0152", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference c3403a4922.\\nC means authorization >= 23.\\nA means condition >= 78.\\nD means restriction >= 71.\\nB means insurance >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 87, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 79}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c7e006\": \"Outcome PROCEED.\", \"candidate_8404c7\": \"Outcome REVIEW.\", \"candidate_6751ef\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c7e006\": 0.13, \"candidate_8404c7\": 0.0, \"candidate_6751ef\": 0.87}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0153:choice", "group_id": "wf12:archive_loan:test:0153", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference dc2b82aa7d.\\nB means insurance >= 45.\\nA means condition >= 77.\\nD means restriction >= 57.\\nC means authorization >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 67, \\\"authorization\\\": 68, \\\"restriction\\\": 46}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1ab0a2\": \"Outcome PROCEED.\", \"candidate_d32a69\": \"Outcome REVIEW.\", \"candidate_f911dc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1ab0a2\": 1.0, \"candidate_d32a69\": 0.0, \"candidate_f911dc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0153:score", "group_id": "wf12:archive_loan:test:0153", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference dc2b82aa7d.\\nB means insurance >= 45.\\nA means condition >= 77.\\nD means restriction >= 57.\\nC means authorization >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 67, \\\"authorization\\\": 68, \\\"restriction\\\": 46}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0153:noul", "group_id": "wf12:archive_loan:test:0153", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference dc2b82aa7d.\\nB means insurance >= 45.\\nA means condition >= 77.\\nD means restriction >= 57.\\nC means authorization >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 67, \\\"authorization\\\": 68, \\\"restriction\\\": 46}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0153:uncertain", "group_id": "wf12:archive_loan:test:0153", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference dc2b82aa7d.\\nB means insurance >= 45.\\nA means condition >= 77.\\nD means restriction >= 57.\\nC means authorization >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 67, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.5369863013698629, \"1\": 0.16205479452054797, \"2\": 0.3009589041095891}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0154:choice", "group_id": "wf12:archive_loan:test:0154", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference aa43538cbd.\\nA means condition >= 90.\\nD means restriction >= 42.\\nB means insurance >= 52.\\nC means authorization >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": 61, \\\"authorization\\\": 40, \\\"restriction\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d552f1\": \"Outcome PROCEED.\", \"candidate_f1e3ec\": \"Outcome REVIEW.\", \"candidate_3bac23\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d552f1\": 0.0, \"candidate_f1e3ec\": 1.0, \"candidate_3bac23\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0154:score", "group_id": "wf12:archive_loan:test:0154", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference aa43538cbd.\\nA means condition >= 90.\\nD means restriction >= 42.\\nB means insurance >= 52.\\nC means authorization >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": 61, \\\"authorization\\\": 40, \\\"restriction\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0154:noul", "group_id": "wf12:archive_loan:test:0154", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference aa43538cbd.\\nA means condition >= 90.\\nD means restriction >= 42.\\nB means insurance >= 52.\\nC means authorization >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": 61, \\\"authorization\\\": 40, \\\"restriction\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0154:uncertain", "group_id": "wf12:archive_loan:test:0154", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference aa43538cbd.\\nA means condition >= 90.\\nD means restriction >= 42.\\nB means insurance >= 52.\\nC means authorization >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 30, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 37}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.9, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.045000000000000005, \"1\": 0.955, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0155:choice", "group_id": "wf12:archive_loan:test:0155", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ff8ef4ce79.\\nB means insurance >= 17.\\nA means condition >= 23.\\nD means restriction >= 25.\\nC means authorization >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 6, \\\"insurance\\\": 12, \\\"authorization\\\": 44, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7e4a89\": \"Outcome PROCEED.\", \"candidate_e93ff5\": \"Outcome REVIEW.\", \"candidate_65352e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7e4a89\": 0.0, \"candidate_e93ff5\": 0.0, \"candidate_65352e\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0155:score", "group_id": "wf12:archive_loan:test:0155", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ff8ef4ce79.\\nB means insurance >= 17.\\nA means condition >= 23.\\nD means restriction >= 25.\\nC means authorization >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 6, \\\"insurance\\\": 12, \\\"authorization\\\": 44, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0155:noul", "group_id": "wf12:archive_loan:test:0155", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ff8ef4ce79.\\nB means insurance >= 17.\\nA means condition >= 23.\\nD means restriction >= 25.\\nC means authorization >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 6, \\\"insurance\\\": 12, \\\"authorization\\\": 44, \\\"restriction\\\": 30}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0155:uncertain", "group_id": "wf12:archive_loan:test:0155", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ff8ef4ce79.\\nB means insurance >= 17.\\nA means condition >= 23.\\nD means restriction >= 25.\\nC means authorization >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": 44, \\\"restriction\\\": 30}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.0, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0156:choice", "group_id": "wf12:archive_loan:test:0156", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 32c5f0023c.\\nB means insurance >= 53.\\nA means condition >= 57.\\nC means authorization >= 90.\\nD means restriction >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 98, \\\"insurance\\\": 76, \\\"authorization\\\": 92, \\\"restriction\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d273bd\": \"Outcome PROCEED.\", \"candidate_f4e8f6\": \"Outcome REVIEW.\", \"candidate_37f0b4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d273bd\": 1.0, \"candidate_f4e8f6\": 0.0, \"candidate_37f0b4\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0156:score", "group_id": "wf12:archive_loan:test:0156", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 32c5f0023c.\\nB means insurance >= 53.\\nA means condition >= 57.\\nC means authorization >= 90.\\nD means restriction >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 98, \\\"insurance\\\": 76, \\\"authorization\\\": 92, \\\"restriction\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0156:noul", "group_id": "wf12:archive_loan:test:0156", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 32c5f0023c.\\nB means insurance >= 53.\\nA means condition >= 57.\\nC means authorization >= 90.\\nD means restriction >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 98, \\\"insurance\\\": 76, \\\"authorization\\\": 92, \\\"restriction\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0156:uncertain", "group_id": "wf12:archive_loan:test:0156", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 32c5f0023c.\\nB means insurance >= 53.\\nA means condition >= 57.\\nC means authorization >= 90.\\nD means restriction >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65, \\\"C\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.904, \"true\": 0.09599999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0157:choice", "group_id": "wf12:archive_loan:test:0157", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ac9472f310.\\nD means restriction >= 37.\\nA means condition >= 83.\\nB means insurance >= 10.\\nC means authorization >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 7, \\\"authorization\\\": 93, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ade36e\": \"Outcome PROCEED.\", \"candidate_6edb47\": \"Outcome REVIEW.\", \"candidate_ab2aab\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ade36e\": 0.0, \"candidate_6edb47\": 1.0, \"candidate_ab2aab\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0157:score", "group_id": "wf12:archive_loan:test:0157", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ac9472f310.\\nD means restriction >= 37.\\nA means condition >= 83.\\nB means insurance >= 10.\\nC means authorization >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 7, \\\"authorization\\\": 93, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0157:noul", "group_id": "wf12:archive_loan:test:0157", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ac9472f310.\\nD means restriction >= 37.\\nA means condition >= 83.\\nB means insurance >= 10.\\nC means authorization >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 79, \\\"insurance\\\": 7, \\\"authorization\\\": 93, \\\"restriction\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0157:uncertain", "group_id": "wf12:archive_loan:test:0157", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference ac9472f310.\\nD means restriction >= 37.\\nA means condition >= 83.\\nB means insurance >= 10.\\nC means authorization >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 7, \\\"authorization\\\": 93, \\\"restriction\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09999999999999998, \"true\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0158:choice", "group_id": "wf12:archive_loan:test:0158", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 98e756357d.\\nB means insurance >= 88.\\nC means authorization >= 56.\\nD means restriction >= 65.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 34, \\\"authorization\\\": 69, \\\"restriction\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_be0446\": \"Outcome PROCEED.\", \"candidate_d714ef\": \"Outcome REVIEW.\", \"candidate_b29768\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_be0446\": 0.0, \"candidate_d714ef\": 0.0, \"candidate_b29768\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0158:score", "group_id": "wf12:archive_loan:test:0158", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 98e756357d.\\nB means insurance >= 88.\\nC means authorization >= 56.\\nD means restriction >= 65.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 34, \\\"authorization\\\": 69, \\\"restriction\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0158:noul", "group_id": "wf12:archive_loan:test:0158", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 98e756357d.\\nB means insurance >= 88.\\nC means authorization >= 56.\\nD means restriction >= 65.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": 34, \\\"authorization\\\": 69, \\\"restriction\\\": 93}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0158:uncertain", "group_id": "wf12:archive_loan:test:0158", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference 98e756357d.\\nB means insurance >= 88.\\nC means authorization >= 56.\\nD means restriction >= 65.\\nA means condition >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 92, \\\"insurance\\\": \\\"unobserved\\\", \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 93}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"C\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.15999999999999992, \"true\": 0.8400000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0159:choice", "group_id": "wf12:archive_loan:test:0159", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a132c46f26.\\nD means restriction >= 66.\\nB means insurance >= 67.\\nA means condition >= 64.\\nC means authorization >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 44, \\\"insurance\\\": 87, \\\"authorization\\\": 93, \\\"restriction\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3d9bb6\": \"Outcome PROCEED.\", \"candidate_ee57cf\": \"Outcome REVIEW.\", \"candidate_48f7f0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3d9bb6\": 1.0, \"candidate_ee57cf\": 0.0, \"candidate_48f7f0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0159:score", "group_id": "wf12:archive_loan:test:0159", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a132c46f26.\\nD means restriction >= 66.\\nB means insurance >= 67.\\nA means condition >= 64.\\nC means authorization >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 44, \\\"insurance\\\": 87, \\\"authorization\\\": 93, \\\"restriction\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0159:noul", "group_id": "wf12:archive_loan:test:0159", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a132c46f26.\\nD means restriction >= 66.\\nB means insurance >= 67.\\nA means condition >= 64.\\nC means authorization >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": 44, \\\"insurance\\\": 87, \\\"authorization\\\": 93, \\\"restriction\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:archive_loan:test:0159:uncertain", "group_id": "wf12:archive_loan:test:0159", "split": "test", "family": "archive_loan", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: archive loan. Case reference a132c46f26.\\nD means restriction >= 66.\\nB means insurance >= 67.\\nA means condition >= 64.\\nC means authorization >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (D AND (NOT B OR NOT C)) THEN HOLD.\\n2. IF (C AND (A OR B)) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"condition\\\": \\\"unobserved\\\", \\\"insurance\\\": 87, \\\"authorization\\\": \\\"unobserved\\\", \\\"restriction\\\": 12}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3d9bb6\": \"Outcome PROCEED.\", \"candidate_ee57cf\": \"Outcome REVIEW.\", \"candidate_48f7f0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3d9bb6\": 0.35, \"candidate_ee57cf\": 0.65, \"candidate_48f7f0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0000:choice", "group_id": "wf12:utility_maintenance:test:0000", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e2531e848d.\\nC means permit >= 76.\\nD means emergency >= 31.\\nB means crew >= 28.\\nA means isolation >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 61, \\\"crew\\\": 81, \\\"permit\\\": 82, \\\"emergency\\\": 25}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f95ecc\": \"Outcome PROCEED.\", \"candidate_5cd7e5\": \"Outcome REVIEW.\", \"candidate_120c89\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f95ecc\": 1.0, \"candidate_5cd7e5\": 0.0, \"candidate_120c89\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0000:score", "group_id": "wf12:utility_maintenance:test:0000", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e2531e848d.\\nC means permit >= 76.\\nD means emergency >= 31.\\nB means crew >= 28.\\nA means isolation >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 61, \\\"crew\\\": 81, \\\"permit\\\": 82, \\\"emergency\\\": 25}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0000:noul", "group_id": "wf12:utility_maintenance:test:0000", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e2531e848d.\\nC means permit >= 76.\\nD means emergency >= 31.\\nB means crew >= 28.\\nA means isolation >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 61, \\\"crew\\\": 81, \\\"permit\\\": 82, \\\"emergency\\\": 25}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0000:uncertain", "group_id": "wf12:utility_maintenance:test:0000", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e2531e848d.\\nC means permit >= 76.\\nD means emergency >= 31.\\nB means crew >= 28.\\nA means isolation >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 81, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 25}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.995, \"true\": 0.005000000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0001:choice", "group_id": "wf12:utility_maintenance:test:0001", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ec41b87fee.\\nC means permit >= 83.\\nD means emergency >= 36.\\nB means crew >= 25.\\nA means isolation >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 66, \\\"permit\\\": 89, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7e376f\": \"Outcome PROCEED.\", \"candidate_1ddbd2\": \"Outcome REVIEW.\", \"candidate_031abd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7e376f\": 0.0, \"candidate_1ddbd2\": 1.0, \"candidate_031abd\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0001:score", "group_id": "wf12:utility_maintenance:test:0001", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ec41b87fee.\\nC means permit >= 83.\\nD means emergency >= 36.\\nB means crew >= 25.\\nA means isolation >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 66, \\\"permit\\\": 89, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0001:noul", "group_id": "wf12:utility_maintenance:test:0001", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ec41b87fee.\\nC means permit >= 83.\\nD means emergency >= 36.\\nB means crew >= 25.\\nA means isolation >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 66, \\\"permit\\\": 89, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0001:uncertain", "group_id": "wf12:utility_maintenance:test:0001", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ec41b87fee.\\nC means permit >= 83.\\nD means emergency >= 36.\\nB means crew >= 25.\\nA means isolation >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 66, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.22750000000000004, \"true\": 0.7725}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0002:choice", "group_id": "wf12:utility_maintenance:test:0002", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3b5e65835f.\\nA means isolation >= 55.\\nB means crew >= 89.\\nC means permit >= 87.\\nD means emergency >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 32, \\\"crew\\\": 15, \\\"permit\\\": 45, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b482bf\": \"Outcome PROCEED.\", \"candidate_a7b7a1\": \"Outcome REVIEW.\", \"candidate_4e8057\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b482bf\": 0.0, \"candidate_a7b7a1\": 0.0, \"candidate_4e8057\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0002:score", "group_id": "wf12:utility_maintenance:test:0002", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3b5e65835f.\\nA means isolation >= 55.\\nB means crew >= 89.\\nC means permit >= 87.\\nD means emergency >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 32, \\\"crew\\\": 15, \\\"permit\\\": 45, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0002:noul", "group_id": "wf12:utility_maintenance:test:0002", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3b5e65835f.\\nA means isolation >= 55.\\nB means crew >= 89.\\nC means permit >= 87.\\nD means emergency >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 32, \\\"crew\\\": 15, \\\"permit\\\": 45, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0002:uncertain", "group_id": "wf12:utility_maintenance:test:0002", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3b5e65835f.\\nA means isolation >= 55.\\nB means crew >= 89.\\nC means permit >= 87.\\nD means emergency >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 15, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 7}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.6295890410958903, \"true\": 0.3704109589041097}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0003:choice", "group_id": "wf12:utility_maintenance:test:0003", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a871ad1fdf.\\nD means emergency >= 50.\\nB means crew >= 67.\\nA means isolation >= 32.\\nC means permit >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 74, \\\"crew\\\": 83, \\\"permit\\\": 45, \\\"emergency\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fd1e73\": \"Outcome PROCEED.\", \"candidate_514252\": \"Outcome REVIEW.\", \"candidate_648e83\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fd1e73\": 1.0, \"candidate_514252\": 0.0, \"candidate_648e83\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0003:score", "group_id": "wf12:utility_maintenance:test:0003", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a871ad1fdf.\\nD means emergency >= 50.\\nB means crew >= 67.\\nA means isolation >= 32.\\nC means permit >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 74, \\\"crew\\\": 83, \\\"permit\\\": 45, \\\"emergency\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0003:noul", "group_id": "wf12:utility_maintenance:test:0003", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a871ad1fdf.\\nD means emergency >= 50.\\nB means crew >= 67.\\nA means isolation >= 32.\\nC means permit >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 74, \\\"crew\\\": 83, \\\"permit\\\": 45, \\\"emergency\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0003:uncertain", "group_id": "wf12:utility_maintenance:test:0003", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a871ad1fdf.\\nD means emergency >= 50.\\nB means crew >= 67.\\nA means isolation >= 32.\\nC means permit >= 45.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 45, \\\"emergency\\\": 35}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"B\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fd1e73\": \"Outcome PROCEED.\", \"candidate_514252\": \"Outcome REVIEW.\", \"candidate_648e83\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fd1e73\": 0.855, \"candidate_514252\": 0.14500000000000002, \"candidate_648e83\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0004:choice", "group_id": "wf12:utility_maintenance:test:0004", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e2547cc6a8.\\nC means permit >= 40.\\nA means isolation >= 58.\\nD means emergency >= 52.\\nB means crew >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 70, \\\"crew\\\": 6, \\\"permit\\\": 16, \\\"emergency\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fd4575\": \"Outcome PROCEED.\", \"candidate_9bc7ba\": \"Outcome REVIEW.\", \"candidate_d71236\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fd4575\": 0.0, \"candidate_9bc7ba\": 1.0, \"candidate_d71236\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0004:score", "group_id": "wf12:utility_maintenance:test:0004", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e2547cc6a8.\\nC means permit >= 40.\\nA means isolation >= 58.\\nD means emergency >= 52.\\nB means crew >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 70, \\\"crew\\\": 6, \\\"permit\\\": 16, \\\"emergency\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0004:noul", "group_id": "wf12:utility_maintenance:test:0004", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e2547cc6a8.\\nC means permit >= 40.\\nA means isolation >= 58.\\nD means emergency >= 52.\\nB means crew >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 70, \\\"crew\\\": 6, \\\"permit\\\": 16, \\\"emergency\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0004:uncertain", "group_id": "wf12:utility_maintenance:test:0004", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e2547cc6a8.\\nC means permit >= 40.\\nA means isolation >= 58.\\nD means emergency >= 52.\\nB means crew >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 16, \\\"emergency\\\": 11}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fd4575\": \"Outcome PROCEED.\", \"candidate_9bc7ba\": \"Outcome REVIEW.\", \"candidate_d71236\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fd4575\": 0.06999999999999999, \"candidate_9bc7ba\": 0.27999999999999997, \"candidate_d71236\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0005:choice", "group_id": "wf12:utility_maintenance:test:0005", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference af1273c064.\\nC means permit >= 64.\\nA means isolation >= 26.\\nD means emergency >= 64.\\nB means crew >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 9, \\\"crew\\\": 19, \\\"permit\\\": 83, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a814e2\": \"Outcome PROCEED.\", \"candidate_018a9c\": \"Outcome REVIEW.\", \"candidate_807a2f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a814e2\": 0.0, \"candidate_018a9c\": 0.0, \"candidate_807a2f\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0005:score", "group_id": "wf12:utility_maintenance:test:0005", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference af1273c064.\\nC means permit >= 64.\\nA means isolation >= 26.\\nD means emergency >= 64.\\nB means crew >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 9, \\\"crew\\\": 19, \\\"permit\\\": 83, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0005:noul", "group_id": "wf12:utility_maintenance:test:0005", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference af1273c064.\\nC means permit >= 64.\\nA means isolation >= 26.\\nD means emergency >= 64.\\nB means crew >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 9, \\\"crew\\\": 19, \\\"permit\\\": 83, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0005:uncertain", "group_id": "wf12:utility_maintenance:test:0005", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference af1273c064.\\nC means permit >= 64.\\nA means isolation >= 26.\\nD means emergency >= 64.\\nB means crew >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.8, \\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a814e2\": \"Outcome PROCEED.\", \"candidate_018a9c\": \"Outcome REVIEW.\", \"candidate_807a2f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a814e2\": 0.735728, \"candidate_018a9c\": 0.1487719999999999, \"candidate_807a2f\": 0.11549999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0006:choice", "group_id": "wf12:utility_maintenance:test:0006", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6a566cf2ae.\\nC means permit >= 65.\\nD means emergency >= 65.\\nA means isolation >= 24.\\nB means crew >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 83, \\\"permit\\\": 33, \\\"emergency\\\": 34}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2d7ba1\": \"Outcome PROCEED.\", \"candidate_a0071e\": \"Outcome REVIEW.\", \"candidate_7514ca\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2d7ba1\": 1.0, \"candidate_a0071e\": 0.0, \"candidate_7514ca\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0006:score", "group_id": "wf12:utility_maintenance:test:0006", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6a566cf2ae.\\nC means permit >= 65.\\nD means emergency >= 65.\\nA means isolation >= 24.\\nB means crew >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 83, \\\"permit\\\": 33, \\\"emergency\\\": 34}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0006:noul", "group_id": "wf12:utility_maintenance:test:0006", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6a566cf2ae.\\nC means permit >= 65.\\nD means emergency >= 65.\\nA means isolation >= 24.\\nB means crew >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 83, \\\"permit\\\": 33, \\\"emergency\\\": 34}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0006:uncertain", "group_id": "wf12:utility_maintenance:test:0006", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6a566cf2ae.\\nC means permit >= 65.\\nD means emergency >= 65.\\nA means isolation >= 24.\\nB means crew >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 83, \\\"permit\\\": 33, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9, \"1\": 0.0, \"2\": 0.09999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0007:choice", "group_id": "wf12:utility_maintenance:test:0007", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 0cb2707e64.\\nA means isolation >= 75.\\nD means emergency >= 77.\\nC means permit >= 49.\\nB means crew >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 96, \\\"crew\\\": 2, \\\"permit\\\": 49, \\\"emergency\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a89528\": \"Outcome PROCEED.\", \"candidate_07e9d8\": \"Outcome REVIEW.\", \"candidate_c5e694\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a89528\": 0.0, \"candidate_07e9d8\": 1.0, \"candidate_c5e694\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0007:score", "group_id": "wf12:utility_maintenance:test:0007", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 0cb2707e64.\\nA means isolation >= 75.\\nD means emergency >= 77.\\nC means permit >= 49.\\nB means crew >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 96, \\\"crew\\\": 2, \\\"permit\\\": 49, \\\"emergency\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0007:noul", "group_id": "wf12:utility_maintenance:test:0007", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 0cb2707e64.\\nA means isolation >= 75.\\nD means emergency >= 77.\\nC means permit >= 49.\\nB means crew >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 96, \\\"crew\\\": 2, \\\"permit\\\": 49, \\\"emergency\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0007:uncertain", "group_id": "wf12:utility_maintenance:test:0007", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 0cb2707e64.\\nA means isolation >= 75.\\nD means emergency >= 77.\\nC means permit >= 49.\\nB means crew >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 49, \\\"emergency\\\": 61}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.52, \"1\": 0.48, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0008:choice", "group_id": "wf12:utility_maintenance:test:0008", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e522168b42.\\nC means permit >= 66.\\nD means emergency >= 47.\\nA means isolation >= 38.\\nB means crew >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 18, \\\"crew\\\": 87, \\\"permit\\\": 43, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ad1af3\": \"Outcome PROCEED.\", \"candidate_2fda45\": \"Outcome REVIEW.\", \"candidate_86f0b1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ad1af3\": 0.0, \"candidate_2fda45\": 0.0, \"candidate_86f0b1\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0008:score", "group_id": "wf12:utility_maintenance:test:0008", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e522168b42.\\nC means permit >= 66.\\nD means emergency >= 47.\\nA means isolation >= 38.\\nB means crew >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 18, \\\"crew\\\": 87, \\\"permit\\\": 43, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0008:noul", "group_id": "wf12:utility_maintenance:test:0008", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e522168b42.\\nC means permit >= 66.\\nD means emergency >= 47.\\nA means isolation >= 38.\\nB means crew >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 18, \\\"crew\\\": 87, \\\"permit\\\": 43, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0008:uncertain", "group_id": "wf12:utility_maintenance:test:0008", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e522168b42.\\nC means permit >= 66.\\nD means emergency >= 47.\\nA means isolation >= 38.\\nB means crew >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.8, \\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.192512, \"1\": 0.13548799999999997, \"2\": 0.6719999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0009:choice", "group_id": "wf12:utility_maintenance:test:0009", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c84b003599.\\nD means emergency >= 80.\\nA means isolation >= 72.\\nB means crew >= 49.\\nC means permit >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 86, \\\"permit\\\": 85, \\\"emergency\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_68ffd2\": \"Outcome PROCEED.\", \"candidate_ed7fb8\": \"Outcome REVIEW.\", \"candidate_024a84\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_68ffd2\": 1.0, \"candidate_ed7fb8\": 0.0, \"candidate_024a84\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0009:score", "group_id": "wf12:utility_maintenance:test:0009", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c84b003599.\\nD means emergency >= 80.\\nA means isolation >= 72.\\nB means crew >= 49.\\nC means permit >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 86, \\\"permit\\\": 85, \\\"emergency\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0009:noul", "group_id": "wf12:utility_maintenance:test:0009", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c84b003599.\\nD means emergency >= 80.\\nA means isolation >= 72.\\nB means crew >= 49.\\nC means permit >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 86, \\\"permit\\\": 85, \\\"emergency\\\": 96}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0009:uncertain", "group_id": "wf12:utility_maintenance:test:0009", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c84b003599.\\nD means emergency >= 80.\\nA means isolation >= 72.\\nB means crew >= 49.\\nC means permit >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 86, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 96}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.95, \"true\": 0.050000000000000044}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0010:choice", "group_id": "wf12:utility_maintenance:test:0010", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1390a34d8e.\\nA means isolation >= 78.\\nB means crew >= 78.\\nD means emergency >= 41.\\nC means permit >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 37, \\\"permit\\\": 17, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e9d3a5\": \"Outcome PROCEED.\", \"candidate_6281bd\": \"Outcome REVIEW.\", \"candidate_4c28da\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e9d3a5\": 0.0, \"candidate_6281bd\": 1.0, \"candidate_4c28da\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0010:score", "group_id": "wf12:utility_maintenance:test:0010", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1390a34d8e.\\nA means isolation >= 78.\\nB means crew >= 78.\\nD means emergency >= 41.\\nC means permit >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 37, \\\"permit\\\": 17, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0010:noul", "group_id": "wf12:utility_maintenance:test:0010", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1390a34d8e.\\nA means isolation >= 78.\\nB means crew >= 78.\\nD means emergency >= 41.\\nC means permit >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 37, \\\"permit\\\": 17, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0010:uncertain", "group_id": "wf12:utility_maintenance:test:0010", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1390a34d8e.\\nA means isolation >= 78.\\nB means crew >= 78.\\nD means emergency >= 41.\\nC means permit >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.2, \\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.3819999999999999, \"true\": 0.6180000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0011:choice", "group_id": "wf12:utility_maintenance:test:0011", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6c9fdceac3.\\nA means isolation >= 51.\\nB means crew >= 13.\\nD means emergency >= 11.\\nC means permit >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 12, \\\"permit\\\": 25, \\\"emergency\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_216f8a\": \"Outcome PROCEED.\", \"candidate_f18968\": \"Outcome REVIEW.\", \"candidate_efa79a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_216f8a\": 0.0, \"candidate_f18968\": 0.0, \"candidate_efa79a\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0011:score", "group_id": "wf12:utility_maintenance:test:0011", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6c9fdceac3.\\nA means isolation >= 51.\\nB means crew >= 13.\\nD means emergency >= 11.\\nC means permit >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 12, \\\"permit\\\": 25, \\\"emergency\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0011:noul", "group_id": "wf12:utility_maintenance:test:0011", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6c9fdceac3.\\nA means isolation >= 51.\\nB means crew >= 13.\\nD means emergency >= 11.\\nC means permit >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 12, \\\"permit\\\": 25, \\\"emergency\\\": 1}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0011:uncertain", "group_id": "wf12:utility_maintenance:test:0011", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6c9fdceac3.\\nA means isolation >= 51.\\nB means crew >= 13.\\nD means emergency >= 11.\\nC means permit >= 56.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 12, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 1}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9324000000000001, \"true\": 0.06759999999999992}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0012:choice", "group_id": "wf12:utility_maintenance:test:0012", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7ce9921fd5.\\nA means isolation >= 23.\\nD means emergency >= 89.\\nB means crew >= 54.\\nC means permit >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": 55, \\\"permit\\\": 2, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eb690f\": \"Outcome PROCEED.\", \"candidate_398d91\": \"Outcome REVIEW.\", \"candidate_e7567f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eb690f\": 1.0, \"candidate_398d91\": 0.0, \"candidate_e7567f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0012:score", "group_id": "wf12:utility_maintenance:test:0012", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7ce9921fd5.\\nA means isolation >= 23.\\nD means emergency >= 89.\\nB means crew >= 54.\\nC means permit >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": 55, \\\"permit\\\": 2, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0012:noul", "group_id": "wf12:utility_maintenance:test:0012", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7ce9921fd5.\\nA means isolation >= 23.\\nD means emergency >= 89.\\nB means crew >= 54.\\nC means permit >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": 55, \\\"permit\\\": 2, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0012:uncertain", "group_id": "wf12:utility_maintenance:test:0012", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7ce9921fd5.\\nA means isolation >= 23.\\nD means emergency >= 89.\\nB means crew >= 54.\\nC means permit >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 2, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eb690f\": \"Outcome PROCEED.\", \"candidate_398d91\": \"Outcome REVIEW.\", \"candidate_e7567f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eb690f\": 0.95, \"candidate_398d91\": 0.050000000000000044, \"candidate_e7567f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0013:choice", "group_id": "wf12:utility_maintenance:test:0013", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7a180fe0a6.\\nA means isolation >= 23.\\nD means emergency >= 12.\\nC means permit >= 34.\\nB means crew >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 69, \\\"crew\\\": 3, \\\"permit\\\": 72, \\\"emergency\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1f6d93\": \"Outcome PROCEED.\", \"candidate_19569a\": \"Outcome REVIEW.\", \"candidate_eeb7cd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1f6d93\": 0.0, \"candidate_19569a\": 1.0, \"candidate_eeb7cd\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0013:score", "group_id": "wf12:utility_maintenance:test:0013", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7a180fe0a6.\\nA means isolation >= 23.\\nD means emergency >= 12.\\nC means permit >= 34.\\nB means crew >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 69, \\\"crew\\\": 3, \\\"permit\\\": 72, \\\"emergency\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0013:noul", "group_id": "wf12:utility_maintenance:test:0013", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7a180fe0a6.\\nA means isolation >= 23.\\nD means emergency >= 12.\\nC means permit >= 34.\\nB means crew >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 69, \\\"crew\\\": 3, \\\"permit\\\": 72, \\\"emergency\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0013:uncertain", "group_id": "wf12:utility_maintenance:test:0013", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7a180fe0a6.\\nA means isolation >= 23.\\nD means emergency >= 12.\\nC means permit >= 34.\\nB means crew >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 3, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 10}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1f6d93\": \"Outcome PROCEED.\", \"candidate_19569a\": \"Outcome REVIEW.\", \"candidate_eeb7cd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1f6d93\": 0.0, \"candidate_19569a\": 0.8400000000000002, \"candidate_eeb7cd\": 0.15999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0014:choice", "group_id": "wf12:utility_maintenance:test:0014", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cb0c40eca0.\\nB means crew >= 71.\\nD means emergency >= 56.\\nA means isolation >= 80.\\nC means permit >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 16, \\\"crew\\\": 98, \\\"permit\\\": 3, \\\"emergency\\\": 25}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_310bf5\": \"Outcome PROCEED.\", \"candidate_4a2300\": \"Outcome REVIEW.\", \"candidate_12b530\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_310bf5\": 0.0, \"candidate_4a2300\": 0.0, \"candidate_12b530\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0014:score", "group_id": "wf12:utility_maintenance:test:0014", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cb0c40eca0.\\nB means crew >= 71.\\nD means emergency >= 56.\\nA means isolation >= 80.\\nC means permit >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 16, \\\"crew\\\": 98, \\\"permit\\\": 3, \\\"emergency\\\": 25}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0014:noul", "group_id": "wf12:utility_maintenance:test:0014", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cb0c40eca0.\\nB means crew >= 71.\\nD means emergency >= 56.\\nA means isolation >= 80.\\nC means permit >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 16, \\\"crew\\\": 98, \\\"permit\\\": 3, \\\"emergency\\\": 25}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0014:uncertain", "group_id": "wf12:utility_maintenance:test:0014", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cb0c40eca0.\\nB means crew >= 71.\\nD means emergency >= 56.\\nA means isolation >= 80.\\nC means permit >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 16, \\\"crew\\\": 98, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_310bf5\": \"Outcome PROCEED.\", \"candidate_4a2300\": \"Outcome REVIEW.\", \"candidate_12b530\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_310bf5\": 0.0, \"candidate_4a2300\": 0.1792, \"candidate_12b530\": 0.8208}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0015:choice", "group_id": "wf12:utility_maintenance:test:0015", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 0518c0e9cb.\\nA means isolation >= 57.\\nC means permit >= 59.\\nB means crew >= 64.\\nD means emergency >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 84, \\\"crew\\\": 96, \\\"permit\\\": 61, \\\"emergency\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2f2952\": \"Outcome PROCEED.\", \"candidate_ea9b9e\": \"Outcome REVIEW.\", \"candidate_82241b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2f2952\": 1.0, \"candidate_ea9b9e\": 0.0, \"candidate_82241b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0015:score", "group_id": "wf12:utility_maintenance:test:0015", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 0518c0e9cb.\\nA means isolation >= 57.\\nC means permit >= 59.\\nB means crew >= 64.\\nD means emergency >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 84, \\\"crew\\\": 96, \\\"permit\\\": 61, \\\"emergency\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0015:noul", "group_id": "wf12:utility_maintenance:test:0015", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 0518c0e9cb.\\nA means isolation >= 57.\\nC means permit >= 59.\\nB means crew >= 64.\\nD means emergency >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 84, \\\"crew\\\": 96, \\\"permit\\\": 61, \\\"emergency\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0015:uncertain", "group_id": "wf12:utility_maintenance:test:0015", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 0518c0e9cb.\\nA means isolation >= 57.\\nC means permit >= 59.\\nB means crew >= 64.\\nD means emergency >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 96, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 82}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.05, \"1\": 0.9025, \"2\": 0.04750000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0016:choice", "group_id": "wf12:utility_maintenance:test:0016", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c41fd5281d.\\nA means isolation >= 56.\\nD means emergency >= 75.\\nB means crew >= 58.\\nC means permit >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 23, \\\"crew\\\": 85, \\\"permit\\\": 80, \\\"emergency\\\": 65}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_71ade4\": \"Outcome PROCEED.\", \"candidate_625a23\": \"Outcome REVIEW.\", \"candidate_1afaa5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_71ade4\": 0.0, \"candidate_625a23\": 1.0, \"candidate_1afaa5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0016:score", "group_id": "wf12:utility_maintenance:test:0016", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c41fd5281d.\\nA means isolation >= 56.\\nD means emergency >= 75.\\nB means crew >= 58.\\nC means permit >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 23, \\\"crew\\\": 85, \\\"permit\\\": 80, \\\"emergency\\\": 65}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0016:noul", "group_id": "wf12:utility_maintenance:test:0016", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c41fd5281d.\\nA means isolation >= 56.\\nD means emergency >= 75.\\nB means crew >= 58.\\nC means permit >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 23, \\\"crew\\\": 85, \\\"permit\\\": 80, \\\"emergency\\\": 65}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0016:uncertain", "group_id": "wf12:utility_maintenance:test:0016", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c41fd5281d.\\nA means isolation >= 56.\\nD means emergency >= 75.\\nB means crew >= 58.\\nC means permit >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 85, \\\"permit\\\": 80, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.42250000000000004, \"2\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0017:choice", "group_id": "wf12:utility_maintenance:test:0017", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c3f5716e35.\\nD means emergency >= 81.\\nA means isolation >= 42.\\nB means crew >= 80.\\nC means permit >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 85, \\\"permit\\\": 25, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7e55bf\": \"Outcome PROCEED.\", \"candidate_a2fb9c\": \"Outcome REVIEW.\", \"candidate_a1d42a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7e55bf\": 0.0, \"candidate_a2fb9c\": 0.0, \"candidate_a1d42a\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0017:score", "group_id": "wf12:utility_maintenance:test:0017", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c3f5716e35.\\nD means emergency >= 81.\\nA means isolation >= 42.\\nB means crew >= 80.\\nC means permit >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 85, \\\"permit\\\": 25, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0017:noul", "group_id": "wf12:utility_maintenance:test:0017", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c3f5716e35.\\nD means emergency >= 81.\\nA means isolation >= 42.\\nB means crew >= 80.\\nC means permit >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 85, \\\"permit\\\": 25, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0017:uncertain", "group_id": "wf12:utility_maintenance:test:0017", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c3f5716e35.\\nD means emergency >= 81.\\nA means isolation >= 42.\\nB means crew >= 80.\\nC means permit >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 25, \\\"emergency\\\": 85}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.22749999999999998, \"1\": 0.12249999999999998, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0018:choice", "group_id": "wf12:utility_maintenance:test:0018", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 20bbc33f9c.\\nB means crew >= 72.\\nA means isolation >= 15.\\nC means permit >= 63.\\nD means emergency >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 26, \\\"crew\\\": 75, \\\"permit\\\": 81, \\\"emergency\\\": 73}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8fa41d\": \"Outcome PROCEED.\", \"candidate_84520c\": \"Outcome REVIEW.\", \"candidate_c2bc7c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8fa41d\": 1.0, \"candidate_84520c\": 0.0, \"candidate_c2bc7c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0018:score", "group_id": "wf12:utility_maintenance:test:0018", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 20bbc33f9c.\\nB means crew >= 72.\\nA means isolation >= 15.\\nC means permit >= 63.\\nD means emergency >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 26, \\\"crew\\\": 75, \\\"permit\\\": 81, \\\"emergency\\\": 73}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0018:noul", "group_id": "wf12:utility_maintenance:test:0018", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 20bbc33f9c.\\nB means crew >= 72.\\nA means isolation >= 15.\\nC means permit >= 63.\\nD means emergency >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 26, \\\"crew\\\": 75, \\\"permit\\\": 81, \\\"emergency\\\": 73}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0018:uncertain", "group_id": "wf12:utility_maintenance:test:0018", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 20bbc33f9c.\\nB means crew >= 72.\\nA means isolation >= 15.\\nC means permit >= 63.\\nD means emergency >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.05, \\\"C\\\": 0.9, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.829, \"true\": 0.17099999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0019:choice", "group_id": "wf12:utility_maintenance:test:0019", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d67bd57923.\\nB means crew >= 39.\\nA means isolation >= 84.\\nD means emergency >= 45.\\nC means permit >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 92, \\\"crew\\\": 8, \\\"permit\\\": 3, \\\"emergency\\\": 64}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a4194\": \"Outcome PROCEED.\", \"candidate_b0c41b\": \"Outcome REVIEW.\", \"candidate_55a4b9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a4194\": 0.0, \"candidate_b0c41b\": 1.0, \"candidate_55a4b9\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0019:score", "group_id": "wf12:utility_maintenance:test:0019", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d67bd57923.\\nB means crew >= 39.\\nA means isolation >= 84.\\nD means emergency >= 45.\\nC means permit >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 92, \\\"crew\\\": 8, \\\"permit\\\": 3, \\\"emergency\\\": 64}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0019:noul", "group_id": "wf12:utility_maintenance:test:0019", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d67bd57923.\\nB means crew >= 39.\\nA means isolation >= 84.\\nD means emergency >= 45.\\nC means permit >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 92, \\\"crew\\\": 8, \\\"permit\\\": 3, \\\"emergency\\\": 64}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0019:uncertain", "group_id": "wf12:utility_maintenance:test:0019", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d67bd57923.\\nB means crew >= 39.\\nA means isolation >= 84.\\nD means emergency >= 45.\\nC means permit >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 3, \\\"emergency\\\": 64}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0020:choice", "group_id": "wf12:utility_maintenance:test:0020", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e4b450a802.\\nD means emergency >= 84.\\nC means permit >= 73.\\nA means isolation >= 25.\\nB means crew >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 2, \\\"crew\\\": 99, \\\"permit\\\": 91, \\\"emergency\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f386e6\": \"Outcome PROCEED.\", \"candidate_f59f1b\": \"Outcome REVIEW.\", \"candidate_912c75\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f386e6\": 0.0, \"candidate_f59f1b\": 0.0, \"candidate_912c75\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0020:score", "group_id": "wf12:utility_maintenance:test:0020", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e4b450a802.\\nD means emergency >= 84.\\nC means permit >= 73.\\nA means isolation >= 25.\\nB means crew >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 2, \\\"crew\\\": 99, \\\"permit\\\": 91, \\\"emergency\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0020:noul", "group_id": "wf12:utility_maintenance:test:0020", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e4b450a802.\\nD means emergency >= 84.\\nC means permit >= 73.\\nA means isolation >= 25.\\nB means crew >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 2, \\\"crew\\\": 99, \\\"permit\\\": 91, \\\"emergency\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0020:uncertain", "group_id": "wf12:utility_maintenance:test:0020", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e4b450a802.\\nD means emergency >= 84.\\nC means permit >= 73.\\nA means isolation >= 25.\\nB means crew >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 99, \\\"permit\\\": 91, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.48, \"true\": 0.52}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0021:choice", "group_id": "wf12:utility_maintenance:test:0021", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference bf02b272df.\\nC means permit >= 75.\\nD means emergency >= 81.\\nB means crew >= 85.\\nA means isolation >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 93, \\\"permit\\\": 81, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dad9d8\": \"Outcome PROCEED.\", \"candidate_a26d07\": \"Outcome REVIEW.\", \"candidate_9e3f8b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dad9d8\": 1.0, \"candidate_a26d07\": 0.0, \"candidate_9e3f8b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0021:score", "group_id": "wf12:utility_maintenance:test:0021", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference bf02b272df.\\nC means permit >= 75.\\nD means emergency >= 81.\\nB means crew >= 85.\\nA means isolation >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 93, \\\"permit\\\": 81, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0021:noul", "group_id": "wf12:utility_maintenance:test:0021", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference bf02b272df.\\nC means permit >= 75.\\nD means emergency >= 81.\\nB means crew >= 85.\\nA means isolation >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 93, \\\"permit\\\": 81, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0021:uncertain", "group_id": "wf12:utility_maintenance:test:0021", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference bf02b272df.\\nC means permit >= 75.\\nD means emergency >= 81.\\nB means crew >= 85.\\nA means isolation >= 88.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 93, \\\"permit\\\": 81, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dad9d8\": \"Outcome PROCEED.\", \"candidate_a26d07\": \"Outcome REVIEW.\", \"candidate_9e3f8b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dad9d8\": 0.9, \"candidate_a26d07\": 0.09499999999999997, \"candidate_9e3f8b\": 0.004999999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0022:choice", "group_id": "wf12:utility_maintenance:test:0022", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c8d3c1e162.\\nA means isolation >= 59.\\nD means emergency >= 34.\\nB means crew >= 44.\\nC means permit >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": 22, \\\"permit\\\": 61, \\\"emergency\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f31ab4\": \"Outcome PROCEED.\", \"candidate_c6722d\": \"Outcome REVIEW.\", \"candidate_3a1223\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f31ab4\": 0.0, \"candidate_c6722d\": 1.0, \"candidate_3a1223\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0022:score", "group_id": "wf12:utility_maintenance:test:0022", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c8d3c1e162.\\nA means isolation >= 59.\\nD means emergency >= 34.\\nB means crew >= 44.\\nC means permit >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": 22, \\\"permit\\\": 61, \\\"emergency\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0022:noul", "group_id": "wf12:utility_maintenance:test:0022", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c8d3c1e162.\\nA means isolation >= 59.\\nD means emergency >= 34.\\nB means crew >= 44.\\nC means permit >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": 22, \\\"permit\\\": 61, \\\"emergency\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0022:uncertain", "group_id": "wf12:utility_maintenance:test:0022", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c8d3c1e162.\\nA means isolation >= 59.\\nD means emergency >= 34.\\nB means crew >= 44.\\nC means permit >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 8}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f31ab4\": \"Outcome PROCEED.\", \"candidate_c6722d\": \"Outcome REVIEW.\", \"candidate_3a1223\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f31ab4\": 0.35, \"candidate_c6722d\": 0.65, \"candidate_3a1223\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0023:choice", "group_id": "wf12:utility_maintenance:test:0023", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f3f7b059b9.\\nD means emergency >= 65.\\nA means isolation >= 15.\\nB means crew >= 82.\\nC means permit >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 2, \\\"crew\\\": 54, \\\"permit\\\": 39, \\\"emergency\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_36ae4c\": \"Outcome PROCEED.\", \"candidate_6443c2\": \"Outcome REVIEW.\", \"candidate_1cef9c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_36ae4c\": 0.0, \"candidate_6443c2\": 0.0, \"candidate_1cef9c\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0023:score", "group_id": "wf12:utility_maintenance:test:0023", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f3f7b059b9.\\nD means emergency >= 65.\\nA means isolation >= 15.\\nB means crew >= 82.\\nC means permit >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 2, \\\"crew\\\": 54, \\\"permit\\\": 39, \\\"emergency\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0023:noul", "group_id": "wf12:utility_maintenance:test:0023", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f3f7b059b9.\\nD means emergency >= 65.\\nA means isolation >= 15.\\nB means crew >= 82.\\nC means permit >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 2, \\\"crew\\\": 54, \\\"permit\\\": 39, \\\"emergency\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0023:uncertain", "group_id": "wf12:utility_maintenance:test:0023", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f3f7b059b9.\\nD means emergency >= 65.\\nA means isolation >= 15.\\nB means crew >= 82.\\nC means permit >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.35, \\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_36ae4c\": \"Outcome PROCEED.\", \"candidate_6443c2\": \"Outcome REVIEW.\", \"candidate_1cef9c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_36ae4c\": 0.4761962328767123, \"candidate_6443c2\": 0.2534287671232877, \"candidate_1cef9c\": 0.270375}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0024:choice", "group_id": "wf12:utility_maintenance:test:0024", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b739aa1ac.\\nD means emergency >= 41.\\nC means permit >= 87.\\nA means isolation >= 31.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 35, \\\"crew\\\": 14, \\\"permit\\\": 88, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f3d192\": \"Outcome PROCEED.\", \"candidate_4f6950\": \"Outcome REVIEW.\", \"candidate_087880\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f3d192\": 1.0, \"candidate_4f6950\": 0.0, \"candidate_087880\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0024:score", "group_id": "wf12:utility_maintenance:test:0024", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b739aa1ac.\\nD means emergency >= 41.\\nC means permit >= 87.\\nA means isolation >= 31.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 35, \\\"crew\\\": 14, \\\"permit\\\": 88, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0024:noul", "group_id": "wf12:utility_maintenance:test:0024", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b739aa1ac.\\nD means emergency >= 41.\\nC means permit >= 87.\\nA means isolation >= 31.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 35, \\\"crew\\\": 14, \\\"permit\\\": 88, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0024:uncertain", "group_id": "wf12:utility_maintenance:test:0024", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b739aa1ac.\\nD means emergency >= 41.\\nC means permit >= 87.\\nA means isolation >= 31.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 88, \\\"emergency\\\": 88}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.95, \"1\": 0.0, \"2\": 0.050000000000000044}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0025:choice", "group_id": "wf12:utility_maintenance:test:0025", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e544517545.\\nA means isolation >= 71.\\nC means permit >= 21.\\nD means emergency >= 51.\\nB means crew >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 54, \\\"crew\\\": 86, \\\"permit\\\": 56, \\\"emergency\\\": 36}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a80f2a\": \"Outcome PROCEED.\", \"candidate_55bb53\": \"Outcome REVIEW.\", \"candidate_1bcc1b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a80f2a\": 0.0, \"candidate_55bb53\": 1.0, \"candidate_1bcc1b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0025:score", "group_id": "wf12:utility_maintenance:test:0025", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e544517545.\\nA means isolation >= 71.\\nC means permit >= 21.\\nD means emergency >= 51.\\nB means crew >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 54, \\\"crew\\\": 86, \\\"permit\\\": 56, \\\"emergency\\\": 36}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0025:noul", "group_id": "wf12:utility_maintenance:test:0025", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e544517545.\\nA means isolation >= 71.\\nC means permit >= 21.\\nD means emergency >= 51.\\nB means crew >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 54, \\\"crew\\\": 86, \\\"permit\\\": 56, \\\"emergency\\\": 36}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0025:uncertain", "group_id": "wf12:utility_maintenance:test:0025", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e544517545.\\nA means isolation >= 71.\\nC means permit >= 21.\\nD means emergency >= 51.\\nB means crew >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 86, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 36}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.52, \"2\": 0.12999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0026:choice", "group_id": "wf12:utility_maintenance:test:0026", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 42a06ca564.\\nD means emergency >= 81.\\nA means isolation >= 83.\\nC means permit >= 44.\\nB means crew >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 65, \\\"crew\\\": 87, \\\"permit\\\": 36, \\\"emergency\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9ab6b8\": \"Outcome PROCEED.\", \"candidate_197d0e\": \"Outcome REVIEW.\", \"candidate_5a1b13\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9ab6b8\": 0.0, \"candidate_197d0e\": 0.0, \"candidate_5a1b13\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0026:score", "group_id": "wf12:utility_maintenance:test:0026", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 42a06ca564.\\nD means emergency >= 81.\\nA means isolation >= 83.\\nC means permit >= 44.\\nB means crew >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 65, \\\"crew\\\": 87, \\\"permit\\\": 36, \\\"emergency\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0026:noul", "group_id": "wf12:utility_maintenance:test:0026", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 42a06ca564.\\nD means emergency >= 81.\\nA means isolation >= 83.\\nC means permit >= 44.\\nB means crew >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 65, \\\"crew\\\": 87, \\\"permit\\\": 36, \\\"emergency\\\": 15}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0026:uncertain", "group_id": "wf12:utility_maintenance:test:0026", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 42a06ca564.\\nD means emergency >= 81.\\nA means isolation >= 83.\\nC means permit >= 44.\\nB means crew >= 77.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65, \\\"C\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.5408000000000001, \"1\": 0.12319999999999998, \"2\": 0.33599999999999997}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0027:choice", "group_id": "wf12:utility_maintenance:test:0027", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e53e6c392e.\\nD means emergency >= 76.\\nC means permit >= 25.\\nB means crew >= 41.\\nA means isolation >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 68, \\\"crew\\\": 27, \\\"permit\\\": 92, \\\"emergency\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ea558a\": \"Outcome PROCEED.\", \"candidate_6afaba\": \"Outcome REVIEW.\", \"candidate_5cc574\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ea558a\": 1.0, \"candidate_6afaba\": 0.0, \"candidate_5cc574\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0027:score", "group_id": "wf12:utility_maintenance:test:0027", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e53e6c392e.\\nD means emergency >= 76.\\nC means permit >= 25.\\nB means crew >= 41.\\nA means isolation >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 68, \\\"crew\\\": 27, \\\"permit\\\": 92, \\\"emergency\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0027:noul", "group_id": "wf12:utility_maintenance:test:0027", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e53e6c392e.\\nD means emergency >= 76.\\nC means permit >= 25.\\nB means crew >= 41.\\nA means isolation >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 68, \\\"crew\\\": 27, \\\"permit\\\": 92, \\\"emergency\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0027:uncertain", "group_id": "wf12:utility_maintenance:test:0027", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e53e6c392e.\\nD means emergency >= 76.\\nC means permit >= 25.\\nB means crew >= 41.\\nA means isolation >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 92, \\\"emergency\\\": 89}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09999999999999998, \"true\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0028:choice", "group_id": "wf12:utility_maintenance:test:0028", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a7543f8980.\\nB means crew >= 57.\\nD means emergency >= 30.\\nC means permit >= 85.\\nA means isolation >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 21, \\\"permit\\\": 92, \\\"emergency\\\": 25}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_264781\": \"Outcome PROCEED.\", \"candidate_bec052\": \"Outcome REVIEW.\", \"candidate_2e3912\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_264781\": 0.0, \"candidate_bec052\": 1.0, \"candidate_2e3912\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0028:score", "group_id": "wf12:utility_maintenance:test:0028", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a7543f8980.\\nB means crew >= 57.\\nD means emergency >= 30.\\nC means permit >= 85.\\nA means isolation >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 21, \\\"permit\\\": 92, \\\"emergency\\\": 25}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0028:noul", "group_id": "wf12:utility_maintenance:test:0028", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a7543f8980.\\nB means crew >= 57.\\nD means emergency >= 30.\\nC means permit >= 85.\\nA means isolation >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 21, \\\"permit\\\": 92, \\\"emergency\\\": 25}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0028:uncertain", "group_id": "wf12:utility_maintenance:test:0028", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a7543f8980.\\nB means crew >= 57.\\nD means emergency >= 30.\\nC means permit >= 85.\\nA means isolation >= 12.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 92, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8, \"true\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0029:choice", "group_id": "wf12:utility_maintenance:test:0029", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7956f96243.\\nD means emergency >= 65.\\nA means isolation >= 80.\\nC means permit >= 69.\\nB means crew >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 59, \\\"permit\\\": 72, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_385bea\": \"Outcome PROCEED.\", \"candidate_026cef\": \"Outcome REVIEW.\", \"candidate_bf5079\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_385bea\": 0.0, \"candidate_026cef\": 0.0, \"candidate_bf5079\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0029:score", "group_id": "wf12:utility_maintenance:test:0029", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7956f96243.\\nD means emergency >= 65.\\nA means isolation >= 80.\\nC means permit >= 69.\\nB means crew >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 59, \\\"permit\\\": 72, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0029:noul", "group_id": "wf12:utility_maintenance:test:0029", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7956f96243.\\nD means emergency >= 65.\\nA means isolation >= 80.\\nC means permit >= 69.\\nB means crew >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 59, \\\"permit\\\": 72, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0029:uncertain", "group_id": "wf12:utility_maintenance:test:0029", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7956f96243.\\nD means emergency >= 65.\\nA means isolation >= 80.\\nC means permit >= 69.\\nB means crew >= 30.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.35, \\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.382, \"true\": 0.618}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0030:choice", "group_id": "wf12:utility_maintenance:test:0030", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ee824fc87d.\\nD means emergency >= 46.\\nB means crew >= 35.\\nA means isolation >= 12.\\nC means permit >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 84, \\\"permit\\\": 79, \\\"emergency\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fe24c0\": \"Outcome PROCEED.\", \"candidate_2f6245\": \"Outcome REVIEW.\", \"candidate_dbddda\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fe24c0\": 1.0, \"candidate_2f6245\": 0.0, \"candidate_dbddda\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0030:score", "group_id": "wf12:utility_maintenance:test:0030", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ee824fc87d.\\nD means emergency >= 46.\\nB means crew >= 35.\\nA means isolation >= 12.\\nC means permit >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 84, \\\"permit\\\": 79, \\\"emergency\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0030:noul", "group_id": "wf12:utility_maintenance:test:0030", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ee824fc87d.\\nD means emergency >= 46.\\nB means crew >= 35.\\nA means isolation >= 12.\\nC means permit >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 84, \\\"permit\\\": 79, \\\"emergency\\\": 27}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0030:uncertain", "group_id": "wf12:utility_maintenance:test:0030", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ee824fc87d.\\nD means emergency >= 46.\\nB means crew >= 35.\\nA means isolation >= 12.\\nC means permit >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 84, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 27}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fe24c0\": \"Outcome PROCEED.\", \"candidate_2f6245\": \"Outcome REVIEW.\", \"candidate_dbddda\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fe24c0\": 0.05, \"candidate_2f6245\": 0.0475, \"candidate_dbddda\": 0.9025}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0031:choice", "group_id": "wf12:utility_maintenance:test:0031", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31a8ed31d8.\\nD means emergency >= 40.\\nA means isolation >= 84.\\nC means permit >= 68.\\nB means crew >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 9, \\\"permit\\\": 80, \\\"emergency\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bbde2a\": \"Outcome PROCEED.\", \"candidate_cbad2b\": \"Outcome REVIEW.\", \"candidate_b5812b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bbde2a\": 0.0, \"candidate_cbad2b\": 1.0, \"candidate_b5812b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0031:score", "group_id": "wf12:utility_maintenance:test:0031", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31a8ed31d8.\\nD means emergency >= 40.\\nA means isolation >= 84.\\nC means permit >= 68.\\nB means crew >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 9, \\\"permit\\\": 80, \\\"emergency\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0031:noul", "group_id": "wf12:utility_maintenance:test:0031", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31a8ed31d8.\\nD means emergency >= 40.\\nA means isolation >= 84.\\nC means permit >= 68.\\nB means crew >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 9, \\\"permit\\\": 80, \\\"emergency\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0031:uncertain", "group_id": "wf12:utility_maintenance:test:0031", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31a8ed31d8.\\nD means emergency >= 40.\\nA means isolation >= 84.\\nC means permit >= 68.\\nB means crew >= 27.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 80, \\\"emergency\\\": 17}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bbde2a\": \"Outcome PROCEED.\", \"candidate_cbad2b\": \"Outcome REVIEW.\", \"candidate_b5812b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bbde2a\": 0.22749999999999998, \"candidate_cbad2b\": 0.7725, \"candidate_b5812b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0032:choice", "group_id": "wf12:utility_maintenance:test:0032", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ece430da89.\\nC means permit >= 25.\\nA means isolation >= 88.\\nB means crew >= 57.\\nD means emergency >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 43, \\\"crew\\\": 0, \\\"permit\\\": 86, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c32058\": \"Outcome PROCEED.\", \"candidate_3db4fb\": \"Outcome REVIEW.\", \"candidate_7e6807\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c32058\": 0.0, \"candidate_3db4fb\": 0.0, \"candidate_7e6807\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0032:score", "group_id": "wf12:utility_maintenance:test:0032", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ece430da89.\\nC means permit >= 25.\\nA means isolation >= 88.\\nB means crew >= 57.\\nD means emergency >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 43, \\\"crew\\\": 0, \\\"permit\\\": 86, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0032:noul", "group_id": "wf12:utility_maintenance:test:0032", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ece430da89.\\nC means permit >= 25.\\nA means isolation >= 88.\\nB means crew >= 57.\\nD means emergency >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 43, \\\"crew\\\": 0, \\\"permit\\\": 86, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0032:uncertain", "group_id": "wf12:utility_maintenance:test:0032", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ece430da89.\\nC means permit >= 25.\\nA means isolation >= 88.\\nB means crew >= 57.\\nD means emergency >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.2, \\\"C\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c32058\": \"Outcome PROCEED.\", \"candidate_3db4fb\": \"Outcome REVIEW.\", \"candidate_7e6807\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c32058\": 0.44720000000000004, \"candidate_3db4fb\": 0.3008, \"candidate_7e6807\": 0.252}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0033:choice", "group_id": "wf12:utility_maintenance:test:0033", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference da42638cfb.\\nC means permit >= 86.\\nA means isolation >= 75.\\nD means emergency >= 25.\\nB means crew >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 60, \\\"permit\\\": 90, \\\"emergency\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2b7fc4\": \"Outcome PROCEED.\", \"candidate_b60947\": \"Outcome REVIEW.\", \"candidate_6a3bf3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2b7fc4\": 1.0, \"candidate_b60947\": 0.0, \"candidate_6a3bf3\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0033:score", "group_id": "wf12:utility_maintenance:test:0033", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference da42638cfb.\\nC means permit >= 86.\\nA means isolation >= 75.\\nD means emergency >= 25.\\nB means crew >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 60, \\\"permit\\\": 90, \\\"emergency\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0033:noul", "group_id": "wf12:utility_maintenance:test:0033", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference da42638cfb.\\nC means permit >= 86.\\nA means isolation >= 75.\\nD means emergency >= 25.\\nB means crew >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 60, \\\"permit\\\": 90, \\\"emergency\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0033:uncertain", "group_id": "wf12:utility_maintenance:test:0033", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference da42638cfb.\\nC means permit >= 86.\\nA means isolation >= 75.\\nD means emergency >= 25.\\nB means crew >= 35.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 50}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.9, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.995, \"1\": 0.005000000000000004, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0034:choice", "group_id": "wf12:utility_maintenance:test:0034", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5002ecfad2.\\nC means permit >= 28.\\nD means emergency >= 40.\\nA means isolation >= 76.\\nB means crew >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 5, \\\"permit\\\": 23, \\\"emergency\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_555c08\": \"Outcome PROCEED.\", \"candidate_c2a488\": \"Outcome REVIEW.\", \"candidate_d93da5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_555c08\": 0.0, \"candidate_c2a488\": 1.0, \"candidate_d93da5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0034:score", "group_id": "wf12:utility_maintenance:test:0034", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5002ecfad2.\\nC means permit >= 28.\\nD means emergency >= 40.\\nA means isolation >= 76.\\nB means crew >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 5, \\\"permit\\\": 23, \\\"emergency\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0034:noul", "group_id": "wf12:utility_maintenance:test:0034", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5002ecfad2.\\nC means permit >= 28.\\nD means emergency >= 40.\\nA means isolation >= 76.\\nB means crew >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 5, \\\"permit\\\": 23, \\\"emergency\\\": 12}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0034:uncertain", "group_id": "wf12:utility_maintenance:test:0034", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5002ecfad2.\\nC means permit >= 28.\\nD means emergency >= 40.\\nA means isolation >= 76.\\nB means crew >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 5, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.42250000000000004, \"1\": 0.5774999999999999, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0035:choice", "group_id": "wf12:utility_maintenance:test:0035", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 229729237c.\\nD means emergency >= 74.\\nC means permit >= 10.\\nA means isolation >= 67.\\nB means crew >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 2, \\\"crew\\\": 56, \\\"permit\\\": 5, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a296d\": \"Outcome PROCEED.\", \"candidate_a96a11\": \"Outcome REVIEW.\", \"candidate_53f315\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a296d\": 0.0, \"candidate_a96a11\": 0.0, \"candidate_53f315\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0035:score", "group_id": "wf12:utility_maintenance:test:0035", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 229729237c.\\nD means emergency >= 74.\\nC means permit >= 10.\\nA means isolation >= 67.\\nB means crew >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 2, \\\"crew\\\": 56, \\\"permit\\\": 5, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0035:noul", "group_id": "wf12:utility_maintenance:test:0035", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 229729237c.\\nD means emergency >= 74.\\nC means permit >= 10.\\nA means isolation >= 67.\\nB means crew >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 2, \\\"crew\\\": 56, \\\"permit\\\": 5, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0035:uncertain", "group_id": "wf12:utility_maintenance:test:0035", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 229729237c.\\nD means emergency >= 74.\\nC means permit >= 10.\\nA means isolation >= 67.\\nB means crew >= 75.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.65, \\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.16910000000000006, \"1\": 0.1289, \"2\": 0.7020000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0036:choice", "group_id": "wf12:utility_maintenance:test:0036", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1c6b94ab88.\\nC means permit >= 26.\\nD means emergency >= 67.\\nA means isolation >= 84.\\nB means crew >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 93, \\\"crew\\\": 69, \\\"permit\\\": 21, \\\"emergency\\\": 73}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_71e157\": \"Outcome PROCEED.\", \"candidate_ff1b6e\": \"Outcome REVIEW.\", \"candidate_450a6e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_71e157\": 1.0, \"candidate_ff1b6e\": 0.0, \"candidate_450a6e\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0036:score", "group_id": "wf12:utility_maintenance:test:0036", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1c6b94ab88.\\nC means permit >= 26.\\nD means emergency >= 67.\\nA means isolation >= 84.\\nB means crew >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 93, \\\"crew\\\": 69, \\\"permit\\\": 21, \\\"emergency\\\": 73}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0036:noul", "group_id": "wf12:utility_maintenance:test:0036", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1c6b94ab88.\\nC means permit >= 26.\\nD means emergency >= 67.\\nA means isolation >= 84.\\nB means crew >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 93, \\\"crew\\\": 69, \\\"permit\\\": 21, \\\"emergency\\\": 73}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0036:uncertain", "group_id": "wf12:utility_maintenance:test:0036", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1c6b94ab88.\\nC means permit >= 26.\\nD means emergency >= 67.\\nA means isolation >= 84.\\nB means crew >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 69, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 73}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.95, \"true\": 0.050000000000000044}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0037:choice", "group_id": "wf12:utility_maintenance:test:0037", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a59ded2627.\\nA means isolation >= 82.\\nD means emergency >= 19.\\nC means permit >= 68.\\nB means crew >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 0, \\\"permit\\\": 16, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f9ac8b\": \"Outcome PROCEED.\", \"candidate_cc56a2\": \"Outcome REVIEW.\", \"candidate_c02d39\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f9ac8b\": 0.0, \"candidate_cc56a2\": 1.0, \"candidate_c02d39\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0037:score", "group_id": "wf12:utility_maintenance:test:0037", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a59ded2627.\\nA means isolation >= 82.\\nD means emergency >= 19.\\nC means permit >= 68.\\nB means crew >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 0, \\\"permit\\\": 16, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0037:noul", "group_id": "wf12:utility_maintenance:test:0037", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a59ded2627.\\nA means isolation >= 82.\\nD means emergency >= 19.\\nC means permit >= 68.\\nB means crew >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 0, \\\"permit\\\": 16, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0037:uncertain", "group_id": "wf12:utility_maintenance:test:0037", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a59ded2627.\\nA means isolation >= 82.\\nD means emergency >= 19.\\nC means permit >= 68.\\nB means crew >= 49.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 0, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 69}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.65, \"true\": 0.35}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0038:choice", "group_id": "wf12:utility_maintenance:test:0038", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7159f8b7b0.\\nB means crew >= 74.\\nD means emergency >= 14.\\nA means isolation >= 73.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": 13, \\\"permit\\\": 87, \\\"emergency\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e23baa\": \"Outcome PROCEED.\", \"candidate_b67345\": \"Outcome REVIEW.\", \"candidate_1c1bf4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e23baa\": 0.0, \"candidate_b67345\": 0.0, \"candidate_1c1bf4\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0038:score", "group_id": "wf12:utility_maintenance:test:0038", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7159f8b7b0.\\nB means crew >= 74.\\nD means emergency >= 14.\\nA means isolation >= 73.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": 13, \\\"permit\\\": 87, \\\"emergency\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0038:noul", "group_id": "wf12:utility_maintenance:test:0038", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7159f8b7b0.\\nB means crew >= 74.\\nD means emergency >= 14.\\nA means isolation >= 73.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": 13, \\\"permit\\\": 87, \\\"emergency\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0038:uncertain", "group_id": "wf12:utility_maintenance:test:0038", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7159f8b7b0.\\nB means crew >= 74.\\nD means emergency >= 14.\\nA means isolation >= 73.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.2, \\\"C\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.904, \"true\": 0.09599999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0039:choice", "group_id": "wf12:utility_maintenance:test:0039", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4a7b0e08ad.\\nB means crew >= 81.\\nD means emergency >= 32.\\nA means isolation >= 77.\\nC means permit >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 92, \\\"permit\\\": 68, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_008dd3\": \"Outcome PROCEED.\", \"candidate_de0058\": \"Outcome REVIEW.\", \"candidate_b69f9c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_008dd3\": 1.0, \"candidate_de0058\": 0.0, \"candidate_b69f9c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0039:score", "group_id": "wf12:utility_maintenance:test:0039", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4a7b0e08ad.\\nB means crew >= 81.\\nD means emergency >= 32.\\nA means isolation >= 77.\\nC means permit >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 92, \\\"permit\\\": 68, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0039:noul", "group_id": "wf12:utility_maintenance:test:0039", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4a7b0e08ad.\\nB means crew >= 81.\\nD means emergency >= 32.\\nA means isolation >= 77.\\nC means permit >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 92, \\\"permit\\\": 68, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0039:uncertain", "group_id": "wf12:utility_maintenance:test:0039", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4a7b0e08ad.\\nB means crew >= 81.\\nD means emergency >= 32.\\nA means isolation >= 77.\\nC means permit >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 57}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_008dd3\": \"Outcome PROCEED.\", \"candidate_de0058\": \"Outcome REVIEW.\", \"candidate_b69f9c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_008dd3\": 0.19000000000000003, \"candidate_de0058\": 0.81, \"candidate_b69f9c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0040:choice", "group_id": "wf12:utility_maintenance:test:0040", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9fa400ad9a.\\nD means emergency >= 10.\\nA means isolation >= 80.\\nB means crew >= 35.\\nC means permit >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 74, \\\"crew\\\": 77, \\\"permit\\\": 66, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2a2c38\": \"Outcome PROCEED.\", \"candidate_145499\": \"Outcome REVIEW.\", \"candidate_56f61f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2a2c38\": 0.0, \"candidate_145499\": 1.0, \"candidate_56f61f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0040:score", "group_id": "wf12:utility_maintenance:test:0040", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9fa400ad9a.\\nD means emergency >= 10.\\nA means isolation >= 80.\\nB means crew >= 35.\\nC means permit >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 74, \\\"crew\\\": 77, \\\"permit\\\": 66, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0040:noul", "group_id": "wf12:utility_maintenance:test:0040", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9fa400ad9a.\\nD means emergency >= 10.\\nA means isolation >= 80.\\nB means crew >= 35.\\nC means permit >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 74, \\\"crew\\\": 77, \\\"permit\\\": 66, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0040:uncertain", "group_id": "wf12:utility_maintenance:test:0040", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9fa400ad9a.\\nD means emergency >= 10.\\nA means isolation >= 80.\\nB means crew >= 35.\\nC means permit >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 74, \\\"crew\\\": 77, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2a2c38\": \"Outcome PROCEED.\", \"candidate_145499\": \"Outcome REVIEW.\", \"candidate_56f61f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2a2c38\": 0.0, \"candidate_145499\": 0.15999999999999998, \"candidate_56f61f\": 0.8400000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0041:choice", "group_id": "wf12:utility_maintenance:test:0041", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e7b42242e0.\\nB means crew >= 37.\\nA means isolation >= 57.\\nC means permit >= 62.\\nD means emergency >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 50, \\\"permit\\\": 88, \\\"emergency\\\": 65}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_92a2e7\": \"Outcome PROCEED.\", \"candidate_7acd68\": \"Outcome REVIEW.\", \"candidate_3d1d7c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_92a2e7\": 0.0, \"candidate_7acd68\": 0.0, \"candidate_3d1d7c\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0041:score", "group_id": "wf12:utility_maintenance:test:0041", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e7b42242e0.\\nB means crew >= 37.\\nA means isolation >= 57.\\nC means permit >= 62.\\nD means emergency >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 50, \\\"permit\\\": 88, \\\"emergency\\\": 65}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0041:noul", "group_id": "wf12:utility_maintenance:test:0041", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e7b42242e0.\\nB means crew >= 37.\\nA means isolation >= 57.\\nC means permit >= 62.\\nD means emergency >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 50, \\\"permit\\\": 88, \\\"emergency\\\": 65}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0041:uncertain", "group_id": "wf12:utility_maintenance:test:0041", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e7b42242e0.\\nB means crew >= 37.\\nA means isolation >= 57.\\nC means permit >= 62.\\nD means emergency >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 88, \\\"emergency\\\": 65}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_92a2e7\": \"Outcome PROCEED.\", \"candidate_7acd68\": \"Outcome REVIEW.\", \"candidate_3d1d7c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_92a2e7\": 0.8, \"candidate_7acd68\": 0.0, \"candidate_3d1d7c\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0042:choice", "group_id": "wf12:utility_maintenance:test:0042", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cfad5c24c9.\\nC means permit >= 15.\\nD means emergency >= 38.\\nB means crew >= 24.\\nA means isolation >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": 10, \\\"permit\\\": 40, \\\"emergency\\\": 45}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_183459\": \"Outcome PROCEED.\", \"candidate_78e5af\": \"Outcome REVIEW.\", \"candidate_26a4fb\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_183459\": 1.0, \"candidate_78e5af\": 0.0, \"candidate_26a4fb\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0042:score", "group_id": "wf12:utility_maintenance:test:0042", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cfad5c24c9.\\nC means permit >= 15.\\nD means emergency >= 38.\\nB means crew >= 24.\\nA means isolation >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": 10, \\\"permit\\\": 40, \\\"emergency\\\": 45}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0042:noul", "group_id": "wf12:utility_maintenance:test:0042", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cfad5c24c9.\\nC means permit >= 15.\\nD means emergency >= 38.\\nB means crew >= 24.\\nA means isolation >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": 10, \\\"permit\\\": 40, \\\"emergency\\\": 45}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0042:uncertain", "group_id": "wf12:utility_maintenance:test:0042", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cfad5c24c9.\\nC means permit >= 15.\\nD means emergency >= 38.\\nB means crew >= 24.\\nA means isolation >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 40, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0975, \"1\": 0.9025, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0043:choice", "group_id": "wf12:utility_maintenance:test:0043", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cc70cd6662.\\nB means crew >= 30.\\nC means permit >= 62.\\nA means isolation >= 58.\\nD means emergency >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 57, \\\"crew\\\": 9, \\\"permit\\\": 99, \\\"emergency\\\": 40}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b71053\": \"Outcome PROCEED.\", \"candidate_8dffd2\": \"Outcome REVIEW.\", \"candidate_3f7717\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b71053\": 0.0, \"candidate_8dffd2\": 1.0, \"candidate_3f7717\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0043:score", "group_id": "wf12:utility_maintenance:test:0043", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cc70cd6662.\\nB means crew >= 30.\\nC means permit >= 62.\\nA means isolation >= 58.\\nD means emergency >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 57, \\\"crew\\\": 9, \\\"permit\\\": 99, \\\"emergency\\\": 40}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0043:noul", "group_id": "wf12:utility_maintenance:test:0043", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cc70cd6662.\\nB means crew >= 30.\\nC means permit >= 62.\\nA means isolation >= 58.\\nD means emergency >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 57, \\\"crew\\\": 9, \\\"permit\\\": 99, \\\"emergency\\\": 40}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0043:uncertain", "group_id": "wf12:utility_maintenance:test:0043", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cc70cd6662.\\nB means crew >= 30.\\nC means permit >= 62.\\nA means isolation >= 58.\\nD means emergency >= 80.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 57, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 99, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.35, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0044:choice", "group_id": "wf12:utility_maintenance:test:0044", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1c8d8b1b4d.\\nC means permit >= 27.\\nD means emergency >= 68.\\nB means crew >= 25.\\nA means isolation >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 53, \\\"crew\\\": 31, \\\"permit\\\": 32, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fc49e7\": \"Outcome PROCEED.\", \"candidate_023fff\": \"Outcome REVIEW.\", \"candidate_63a6a9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fc49e7\": 0.0, \"candidate_023fff\": 0.0, \"candidate_63a6a9\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0044:score", "group_id": "wf12:utility_maintenance:test:0044", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1c8d8b1b4d.\\nC means permit >= 27.\\nD means emergency >= 68.\\nB means crew >= 25.\\nA means isolation >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 53, \\\"crew\\\": 31, \\\"permit\\\": 32, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0044:noul", "group_id": "wf12:utility_maintenance:test:0044", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1c8d8b1b4d.\\nC means permit >= 27.\\nD means emergency >= 68.\\nB means crew >= 25.\\nA means isolation >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 53, \\\"crew\\\": 31, \\\"permit\\\": 32, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0044:uncertain", "group_id": "wf12:utility_maintenance:test:0044", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1c8d8b1b4d.\\nC means permit >= 27.\\nD means emergency >= 68.\\nB means crew >= 25.\\nA means isolation >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 32, \\\"emergency\\\": 88}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.5369863013698629, \"1\": 0.0, \"2\": 0.4630136986301371}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0045:choice", "group_id": "wf12:utility_maintenance:test:0045", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 90711ddb11.\\nA means isolation >= 34.\\nD means emergency >= 21.\\nC means permit >= 69.\\nB means crew >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": 28, \\\"permit\\\": 74, \\\"emergency\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bcec6a\": \"Outcome PROCEED.\", \"candidate_c635bb\": \"Outcome REVIEW.\", \"candidate_1cbe13\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bcec6a\": 1.0, \"candidate_c635bb\": 0.0, \"candidate_1cbe13\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0045:score", "group_id": "wf12:utility_maintenance:test:0045", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 90711ddb11.\\nA means isolation >= 34.\\nD means emergency >= 21.\\nC means permit >= 69.\\nB means crew >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": 28, \\\"permit\\\": 74, \\\"emergency\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0045:noul", "group_id": "wf12:utility_maintenance:test:0045", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 90711ddb11.\\nA means isolation >= 34.\\nD means emergency >= 21.\\nC means permit >= 69.\\nB means crew >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": 28, \\\"permit\\\": 74, \\\"emergency\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0045:uncertain", "group_id": "wf12:utility_maintenance:test:0045", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 90711ddb11.\\nA means isolation >= 34.\\nD means emergency >= 21.\\nC means permit >= 69.\\nB means crew >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"B\\\": 0.1, \\\"C\\\": 0.95, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.90475, \"true\": 0.09524999999999997}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0046:choice", "group_id": "wf12:utility_maintenance:test:0046", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a74b420d15.\\nB means crew >= 69.\\nA means isolation >= 40.\\nD means emergency >= 32.\\nC means permit >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 89, \\\"permit\\\": 83, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b1b641\": \"Outcome PROCEED.\", \"candidate_0ff23a\": \"Outcome REVIEW.\", \"candidate_8bc805\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b1b641\": 0.0, \"candidate_0ff23a\": 1.0, \"candidate_8bc805\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0046:score", "group_id": "wf12:utility_maintenance:test:0046", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a74b420d15.\\nB means crew >= 69.\\nA means isolation >= 40.\\nD means emergency >= 32.\\nC means permit >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 89, \\\"permit\\\": 83, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0046:noul", "group_id": "wf12:utility_maintenance:test:0046", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a74b420d15.\\nB means crew >= 69.\\nA means isolation >= 40.\\nD means emergency >= 32.\\nC means permit >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 89, \\\"permit\\\": 83, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0046:uncertain", "group_id": "wf12:utility_maintenance:test:0046", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a74b420d15.\\nB means crew >= 69.\\nA means isolation >= 40.\\nD means emergency >= 32.\\nC means permit >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.8, \\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.2979999999999998, \"true\": 0.7020000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0047:choice", "group_id": "wf12:utility_maintenance:test:0047", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9a858190f0.\\nA means isolation >= 79.\\nD means emergency >= 51.\\nC means permit >= 10.\\nB means crew >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 3, \\\"permit\\\": 2, \\\"emergency\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f8a0b1\": \"Outcome PROCEED.\", \"candidate_ec9b47\": \"Outcome REVIEW.\", \"candidate_02b8c2\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f8a0b1\": 0.0, \"candidate_ec9b47\": 0.0, \"candidate_02b8c2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0047:score", "group_id": "wf12:utility_maintenance:test:0047", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9a858190f0.\\nA means isolation >= 79.\\nD means emergency >= 51.\\nC means permit >= 10.\\nB means crew >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 3, \\\"permit\\\": 2, \\\"emergency\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0047:noul", "group_id": "wf12:utility_maintenance:test:0047", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9a858190f0.\\nA means isolation >= 79.\\nD means emergency >= 51.\\nC means permit >= 10.\\nB means crew >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 3, \\\"permit\\\": 2, \\\"emergency\\\": 33}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0047:uncertain", "group_id": "wf12:utility_maintenance:test:0047", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9a858190f0.\\nA means isolation >= 79.\\nD means emergency >= 51.\\nC means permit >= 10.\\nB means crew >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 3, \\\"permit\\\": 2, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.5369863013698629, \"true\": 0.4630136986301371}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0048:choice", "group_id": "wf12:utility_maintenance:test:0048", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference de6755267b.\\nA means isolation >= 12.\\nD means emergency >= 32.\\nB means crew >= 30.\\nC means permit >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 18, \\\"crew\\\": 32, \\\"permit\\\": 98, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_98c349\": \"Outcome PROCEED.\", \"candidate_aa22a6\": \"Outcome REVIEW.\", \"candidate_3fb1e0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_98c349\": 1.0, \"candidate_aa22a6\": 0.0, \"candidate_3fb1e0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0048:score", "group_id": "wf12:utility_maintenance:test:0048", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference de6755267b.\\nA means isolation >= 12.\\nD means emergency >= 32.\\nB means crew >= 30.\\nC means permit >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 18, \\\"crew\\\": 32, \\\"permit\\\": 98, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0048:noul", "group_id": "wf12:utility_maintenance:test:0048", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference de6755267b.\\nA means isolation >= 12.\\nD means emergency >= 32.\\nB means crew >= 30.\\nC means permit >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 18, \\\"crew\\\": 32, \\\"permit\\\": 98, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0048:uncertain", "group_id": "wf12:utility_maintenance:test:0048", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference de6755267b.\\nA means isolation >= 12.\\nD means emergency >= 32.\\nB means crew >= 30.\\nC means permit >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 18, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 7}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.9, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_98c349\": \"Outcome PROCEED.\", \"candidate_aa22a6\": \"Outcome REVIEW.\", \"candidate_3fb1e0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_98c349\": 0.9, \"candidate_aa22a6\": 0.09999999999999998, \"candidate_3fb1e0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0049:choice", "group_id": "wf12:utility_maintenance:test:0049", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 06a58abb82.\\nA means isolation >= 56.\\nD means emergency >= 83.\\nB means crew >= 82.\\nC means permit >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 25, \\\"crew\\\": 98, \\\"permit\\\": 89, \\\"emergency\\\": 70}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_27aa39\": \"Outcome PROCEED.\", \"candidate_34f1ed\": \"Outcome REVIEW.\", \"candidate_148f36\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_27aa39\": 0.0, \"candidate_34f1ed\": 1.0, \"candidate_148f36\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0049:score", "group_id": "wf12:utility_maintenance:test:0049", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 06a58abb82.\\nA means isolation >= 56.\\nD means emergency >= 83.\\nB means crew >= 82.\\nC means permit >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 25, \\\"crew\\\": 98, \\\"permit\\\": 89, \\\"emergency\\\": 70}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0049:noul", "group_id": "wf12:utility_maintenance:test:0049", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 06a58abb82.\\nA means isolation >= 56.\\nD means emergency >= 83.\\nB means crew >= 82.\\nC means permit >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 25, \\\"crew\\\": 98, \\\"permit\\\": 89, \\\"emergency\\\": 70}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0049:uncertain", "group_id": "wf12:utility_maintenance:test:0049", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 06a58abb82.\\nA means isolation >= 56.\\nD means emergency >= 83.\\nB means crew >= 82.\\nC means permit >= 68.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 98, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 70}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_27aa39\": \"Outcome PROCEED.\", \"candidate_34f1ed\": \"Outcome REVIEW.\", \"candidate_148f36\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_27aa39\": 0.35, \"candidate_34f1ed\": 0.42250000000000004, \"candidate_148f36\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0050:choice", "group_id": "wf12:utility_maintenance:test:0050", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 046a783b69.\\nC means permit >= 38.\\nB means crew >= 27.\\nD means emergency >= 60.\\nA means isolation >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 41, \\\"crew\\\": 1, \\\"permit\\\": 53, \\\"emergency\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2deaf9\": \"Outcome PROCEED.\", \"candidate_6f3fb8\": \"Outcome REVIEW.\", \"candidate_a8250f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2deaf9\": 0.0, \"candidate_6f3fb8\": 0.0, \"candidate_a8250f\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0050:score", "group_id": "wf12:utility_maintenance:test:0050", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 046a783b69.\\nC means permit >= 38.\\nB means crew >= 27.\\nD means emergency >= 60.\\nA means isolation >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 41, \\\"crew\\\": 1, \\\"permit\\\": 53, \\\"emergency\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0050:noul", "group_id": "wf12:utility_maintenance:test:0050", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 046a783b69.\\nC means permit >= 38.\\nB means crew >= 27.\\nD means emergency >= 60.\\nA means isolation >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 41, \\\"crew\\\": 1, \\\"permit\\\": 53, \\\"emergency\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0050:uncertain", "group_id": "wf12:utility_maintenance:test:0050", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 046a783b69.\\nC means permit >= 38.\\nB means crew >= 27.\\nD means emergency >= 60.\\nA means isolation >= 55.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 41, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 53, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2deaf9\": \"Outcome PROCEED.\", \"candidate_6f3fb8\": \"Outcome REVIEW.\", \"candidate_a8250f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2deaf9\": 0.0, \"candidate_6f3fb8\": 0.35, \"candidate_a8250f\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0051:choice", "group_id": "wf12:utility_maintenance:test:0051", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ed9806de2c.\\nB means crew >= 42.\\nA means isolation >= 30.\\nC means permit >= 70.\\nD means emergency >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 49, \\\"crew\\\": 63, \\\"permit\\\": 20, \\\"emergency\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5ee3b9\": \"Outcome PROCEED.\", \"candidate_c9e192\": \"Outcome REVIEW.\", \"candidate_4568ad\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5ee3b9\": 1.0, \"candidate_c9e192\": 0.0, \"candidate_4568ad\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0051:score", "group_id": "wf12:utility_maintenance:test:0051", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ed9806de2c.\\nB means crew >= 42.\\nA means isolation >= 30.\\nC means permit >= 70.\\nD means emergency >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 49, \\\"crew\\\": 63, \\\"permit\\\": 20, \\\"emergency\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0051:noul", "group_id": "wf12:utility_maintenance:test:0051", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ed9806de2c.\\nB means crew >= 42.\\nA means isolation >= 30.\\nC means permit >= 70.\\nD means emergency >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 49, \\\"crew\\\": 63, \\\"permit\\\": 20, \\\"emergency\\\": 90}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0051:uncertain", "group_id": "wf12:utility_maintenance:test:0051", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ed9806de2c.\\nB means crew >= 42.\\nA means isolation >= 30.\\nC means permit >= 70.\\nD means emergency >= 81.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 20, \\\"emergency\\\": 90}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.005000000000000001, \"1\": 0.095, \"2\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0052:choice", "group_id": "wf12:utility_maintenance:test:0052", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 24a9a8d1b8.\\nC means permit >= 21.\\nD means emergency >= 67.\\nB means crew >= 52.\\nA means isolation >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": 48, \\\"permit\\\": 37, \\\"emergency\\\": 24}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cc296f\": \"Outcome PROCEED.\", \"candidate_aa5e61\": \"Outcome REVIEW.\", \"candidate_4020cd\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cc296f\": 0.0, \"candidate_aa5e61\": 1.0, \"candidate_4020cd\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0052:score", "group_id": "wf12:utility_maintenance:test:0052", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 24a9a8d1b8.\\nC means permit >= 21.\\nD means emergency >= 67.\\nB means crew >= 52.\\nA means isolation >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": 48, \\\"permit\\\": 37, \\\"emergency\\\": 24}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0052:noul", "group_id": "wf12:utility_maintenance:test:0052", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 24a9a8d1b8.\\nC means permit >= 21.\\nD means emergency >= 67.\\nB means crew >= 52.\\nA means isolation >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": 48, \\\"permit\\\": 37, \\\"emergency\\\": 24}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0052:uncertain", "group_id": "wf12:utility_maintenance:test:0052", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 24a9a8d1b8.\\nC means permit >= 21.\\nD means emergency >= 67.\\nB means crew >= 52.\\nA means isolation >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 37, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.7725, \"1\": 0.22749999999999998, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0053:choice", "group_id": "wf12:utility_maintenance:test:0053", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 57d2d22c84.\\nB means crew >= 65.\\nA means isolation >= 46.\\nD means emergency >= 35.\\nC means permit >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 24, \\\"crew\\\": 77, \\\"permit\\\": 83, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e08b07\": \"Outcome PROCEED.\", \"candidate_194e03\": \"Outcome REVIEW.\", \"candidate_c51a2e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e08b07\": 0.0, \"candidate_194e03\": 0.0, \"candidate_c51a2e\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0053:score", "group_id": "wf12:utility_maintenance:test:0053", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 57d2d22c84.\\nB means crew >= 65.\\nA means isolation >= 46.\\nD means emergency >= 35.\\nC means permit >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 24, \\\"crew\\\": 77, \\\"permit\\\": 83, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0053:noul", "group_id": "wf12:utility_maintenance:test:0053", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 57d2d22c84.\\nB means crew >= 65.\\nA means isolation >= 46.\\nD means emergency >= 35.\\nC means permit >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 24, \\\"crew\\\": 77, \\\"permit\\\": 83, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0053:uncertain", "group_id": "wf12:utility_maintenance:test:0053", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 57d2d22c84.\\nB means crew >= 65.\\nA means isolation >= 46.\\nD means emergency >= 35.\\nC means permit >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 77, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 85}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.0, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0054:choice", "group_id": "wf12:utility_maintenance:test:0054", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f2e7139cde.\\nC means permit >= 24.\\nA means isolation >= 32.\\nB means crew >= 26.\\nD means emergency >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 52, \\\"crew\\\": 28, \\\"permit\\\": 19, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_faaf5d\": \"Outcome PROCEED.\", \"candidate_f62ede\": \"Outcome REVIEW.\", \"candidate_c55292\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_faaf5d\": 1.0, \"candidate_f62ede\": 0.0, \"candidate_c55292\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0054:score", "group_id": "wf12:utility_maintenance:test:0054", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f2e7139cde.\\nC means permit >= 24.\\nA means isolation >= 32.\\nB means crew >= 26.\\nD means emergency >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 52, \\\"crew\\\": 28, \\\"permit\\\": 19, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0054:noul", "group_id": "wf12:utility_maintenance:test:0054", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f2e7139cde.\\nC means permit >= 24.\\nA means isolation >= 32.\\nB means crew >= 26.\\nD means emergency >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 52, \\\"crew\\\": 28, \\\"permit\\\": 19, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0054:uncertain", "group_id": "wf12:utility_maintenance:test:0054", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f2e7139cde.\\nC means permit >= 24.\\nA means isolation >= 32.\\nB means crew >= 26.\\nD means emergency >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"B\\\": 0.95, \\\"C\\\": 0.1, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.909, \"true\": 0.09099999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0055:choice", "group_id": "wf12:utility_maintenance:test:0055", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f7f35456a4.\\nD means emergency >= 58.\\nC means permit >= 70.\\nB means crew >= 23.\\nA means isolation >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 4, \\\"permit\\\": 88, \\\"emergency\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_202ebf\": \"Outcome PROCEED.\", \"candidate_7ca62e\": \"Outcome REVIEW.\", \"candidate_25d4de\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_202ebf\": 0.0, \"candidate_7ca62e\": 1.0, \"candidate_25d4de\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0055:score", "group_id": "wf12:utility_maintenance:test:0055", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f7f35456a4.\\nD means emergency >= 58.\\nC means permit >= 70.\\nB means crew >= 23.\\nA means isolation >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 4, \\\"permit\\\": 88, \\\"emergency\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0055:noul", "group_id": "wf12:utility_maintenance:test:0055", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f7f35456a4.\\nD means emergency >= 58.\\nC means permit >= 70.\\nB means crew >= 23.\\nA means isolation >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 4, \\\"permit\\\": 88, \\\"emergency\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0055:uncertain", "group_id": "wf12:utility_maintenance:test:0055", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f7f35456a4.\\nD means emergency >= 58.\\nC means permit >= 70.\\nB means crew >= 23.\\nA means isolation >= 40.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 35}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0056:choice", "group_id": "wf12:utility_maintenance:test:0056", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 884a47cd36.\\nD means emergency >= 63.\\nA means isolation >= 39.\\nB means crew >= 61.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 37, \\\"crew\\\": 20, \\\"permit\\\": 65, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_598df6\": \"Outcome PROCEED.\", \"candidate_38d13f\": \"Outcome REVIEW.\", \"candidate_962fef\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_598df6\": 0.0, \"candidate_38d13f\": 0.0, \"candidate_962fef\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0056:score", "group_id": "wf12:utility_maintenance:test:0056", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 884a47cd36.\\nD means emergency >= 63.\\nA means isolation >= 39.\\nB means crew >= 61.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 37, \\\"crew\\\": 20, \\\"permit\\\": 65, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0056:noul", "group_id": "wf12:utility_maintenance:test:0056", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 884a47cd36.\\nD means emergency >= 63.\\nA means isolation >= 39.\\nB means crew >= 61.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 37, \\\"crew\\\": 20, \\\"permit\\\": 65, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0056:uncertain", "group_id": "wf12:utility_maintenance:test:0056", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 884a47cd36.\\nD means emergency >= 63.\\nA means isolation >= 39.\\nB means crew >= 61.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.65, \\\"C\\\": 0.8, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.42399999999999993, \"true\": 0.5760000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0057:choice", "group_id": "wf12:utility_maintenance:test:0057", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9c9e349597.\\nC means permit >= 37.\\nB means crew >= 24.\\nA means isolation >= 16.\\nD means emergency >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 17, \\\"crew\\\": 68, \\\"permit\\\": 18, \\\"emergency\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c65a21\": \"Outcome PROCEED.\", \"candidate_fe7803\": \"Outcome REVIEW.\", \"candidate_7898a4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c65a21\": 1.0, \"candidate_fe7803\": 0.0, \"candidate_7898a4\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0057:score", "group_id": "wf12:utility_maintenance:test:0057", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9c9e349597.\\nC means permit >= 37.\\nB means crew >= 24.\\nA means isolation >= 16.\\nD means emergency >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 17, \\\"crew\\\": 68, \\\"permit\\\": 18, \\\"emergency\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0057:noul", "group_id": "wf12:utility_maintenance:test:0057", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9c9e349597.\\nC means permit >= 37.\\nB means crew >= 24.\\nA means isolation >= 16.\\nD means emergency >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 17, \\\"crew\\\": 68, \\\"permit\\\": 18, \\\"emergency\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0057:uncertain", "group_id": "wf12:utility_maintenance:test:0057", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9c9e349597.\\nC means permit >= 37.\\nB means crew >= 24.\\nA means isolation >= 16.\\nD means emergency >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 17, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 18, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c65a21\": \"Outcome PROCEED.\", \"candidate_fe7803\": \"Outcome REVIEW.\", \"candidate_7898a4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c65a21\": 0.95, \"candidate_fe7803\": 0.050000000000000044, \"candidate_7898a4\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0058:choice", "group_id": "wf12:utility_maintenance:test:0058", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31476fd172.\\nD means emergency >= 31.\\nB means crew >= 25.\\nC means permit >= 44.\\nA means isolation >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 5, \\\"permit\\\": 24, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_40908f\": \"Outcome PROCEED.\", \"candidate_67e776\": \"Outcome REVIEW.\", \"candidate_e41be0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_40908f\": 0.0, \"candidate_67e776\": 1.0, \"candidate_e41be0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0058:score", "group_id": "wf12:utility_maintenance:test:0058", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31476fd172.\\nD means emergency >= 31.\\nB means crew >= 25.\\nC means permit >= 44.\\nA means isolation >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 5, \\\"permit\\\": 24, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0058:noul", "group_id": "wf12:utility_maintenance:test:0058", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31476fd172.\\nD means emergency >= 31.\\nB means crew >= 25.\\nC means permit >= 44.\\nA means isolation >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 5, \\\"permit\\\": 24, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0058:uncertain", "group_id": "wf12:utility_maintenance:test:0058", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31476fd172.\\nD means emergency >= 31.\\nB means crew >= 25.\\nC means permit >= 44.\\nA means isolation >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 5, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_40908f\": \"Outcome PROCEED.\", \"candidate_67e776\": \"Outcome REVIEW.\", \"candidate_e41be0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_40908f\": 0.27999999999999997, \"candidate_67e776\": 0.72, \"candidate_e41be0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0059:choice", "group_id": "wf12:utility_maintenance:test:0059", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d2783593ae.\\nC means permit >= 36.\\nA means isolation >= 66.\\nD means emergency >= 41.\\nB means crew >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 48, \\\"crew\\\": 90, \\\"permit\\\": 19, \\\"emergency\\\": 44}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0bfc03\": \"Outcome PROCEED.\", \"candidate_33d643\": \"Outcome REVIEW.\", \"candidate_b4f7d6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0bfc03\": 0.0, \"candidate_33d643\": 0.0, \"candidate_b4f7d6\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0059:score", "group_id": "wf12:utility_maintenance:test:0059", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d2783593ae.\\nC means permit >= 36.\\nA means isolation >= 66.\\nD means emergency >= 41.\\nB means crew >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 48, \\\"crew\\\": 90, \\\"permit\\\": 19, \\\"emergency\\\": 44}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0059:noul", "group_id": "wf12:utility_maintenance:test:0059", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d2783593ae.\\nC means permit >= 36.\\nA means isolation >= 66.\\nD means emergency >= 41.\\nB means crew >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 48, \\\"crew\\\": 90, \\\"permit\\\": 19, \\\"emergency\\\": 44}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0059:uncertain", "group_id": "wf12:utility_maintenance:test:0059", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d2783593ae.\\nC means permit >= 36.\\nA means isolation >= 66.\\nD means emergency >= 41.\\nB means crew >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 90, \\\"permit\\\": 19, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0bfc03\": \"Outcome PROCEED.\", \"candidate_33d643\": \"Outcome REVIEW.\", \"candidate_b4f7d6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0bfc03\": 0.8, \"candidate_33d643\": 0.0, \"candidate_b4f7d6\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0060:choice", "group_id": "wf12:utility_maintenance:test:0060", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ce23400ff7.\\nC means permit >= 68.\\nA means isolation >= 12.\\nD means emergency >= 55.\\nB means crew >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": 15, \\\"permit\\\": 90, \\\"emergency\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_95e7a6\": \"Outcome PROCEED.\", \"candidate_d7116c\": \"Outcome REVIEW.\", \"candidate_f246f7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_95e7a6\": 1.0, \"candidate_d7116c\": 0.0, \"candidate_f246f7\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0060:score", "group_id": "wf12:utility_maintenance:test:0060", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ce23400ff7.\\nC means permit >= 68.\\nA means isolation >= 12.\\nD means emergency >= 55.\\nB means crew >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": 15, \\\"permit\\\": 90, \\\"emergency\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0060:noul", "group_id": "wf12:utility_maintenance:test:0060", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ce23400ff7.\\nC means permit >= 68.\\nA means isolation >= 12.\\nD means emergency >= 55.\\nB means crew >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": 15, \\\"permit\\\": 90, \\\"emergency\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0060:uncertain", "group_id": "wf12:utility_maintenance:test:0060", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ce23400ff7.\\nC means permit >= 68.\\nA means isolation >= 12.\\nD means emergency >= 55.\\nB means crew >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 90, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.955, \"1\": 0.04500000000000004, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0061:choice", "group_id": "wf12:utility_maintenance:test:0061", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 8987b9e14f.\\nA means isolation >= 70.\\nD means emergency >= 17.\\nC means permit >= 75.\\nB means crew >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": 48, \\\"permit\\\": 79, \\\"emergency\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d05634\": \"Outcome PROCEED.\", \"candidate_f7c197\": \"Outcome REVIEW.\", \"candidate_5bfc2a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d05634\": 0.0, \"candidate_f7c197\": 1.0, \"candidate_5bfc2a\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0061:score", "group_id": "wf12:utility_maintenance:test:0061", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 8987b9e14f.\\nA means isolation >= 70.\\nD means emergency >= 17.\\nC means permit >= 75.\\nB means crew >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": 48, \\\"permit\\\": 79, \\\"emergency\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0061:noul", "group_id": "wf12:utility_maintenance:test:0061", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 8987b9e14f.\\nA means isolation >= 70.\\nD means emergency >= 17.\\nC means permit >= 75.\\nB means crew >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": 48, \\\"permit\\\": 79, \\\"emergency\\\": 10}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0061:uncertain", "group_id": "wf12:utility_maintenance:test:0061", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 8987b9e14f.\\nA means isolation >= 70.\\nD means emergency >= 17.\\nC means permit >= 75.\\nB means crew >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 48, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 10}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.8400000000000002, \"2\": 0.15999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0062:choice", "group_id": "wf12:utility_maintenance:test:0062", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f157dd30ac.\\nB means crew >= 88.\\nC means permit >= 25.\\nA means isolation >= 39.\\nD means emergency >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 31, \\\"crew\\\": 88, \\\"permit\\\": 6, \\\"emergency\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e0208c\": \"Outcome PROCEED.\", \"candidate_a223ac\": \"Outcome REVIEW.\", \"candidate_bc76a8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e0208c\": 0.0, \"candidate_a223ac\": 0.0, \"candidate_bc76a8\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0062:score", "group_id": "wf12:utility_maintenance:test:0062", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f157dd30ac.\\nB means crew >= 88.\\nC means permit >= 25.\\nA means isolation >= 39.\\nD means emergency >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 31, \\\"crew\\\": 88, \\\"permit\\\": 6, \\\"emergency\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0062:noul", "group_id": "wf12:utility_maintenance:test:0062", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f157dd30ac.\\nB means crew >= 88.\\nC means permit >= 25.\\nA means isolation >= 39.\\nD means emergency >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 31, \\\"crew\\\": 88, \\\"permit\\\": 6, \\\"emergency\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0062:uncertain", "group_id": "wf12:utility_maintenance:test:0062", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f157dd30ac.\\nB means crew >= 88.\\nC means permit >= 25.\\nA means isolation >= 39.\\nD means emergency >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 88, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 29}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8960000000000001, \"1\": 0.03639999999999995, \"2\": 0.06759999999999992}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0063:choice", "group_id": "wf12:utility_maintenance:test:0063", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 400ef8388a.\\nA means isolation >= 60.\\nD means emergency >= 75.\\nC means permit >= 85.\\nB means crew >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": 88, \\\"permit\\\": 93, \\\"emergency\\\": 47}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_92d070\": \"Outcome PROCEED.\", \"candidate_4ce6ee\": \"Outcome REVIEW.\", \"candidate_016f18\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_92d070\": 1.0, \"candidate_4ce6ee\": 0.0, \"candidate_016f18\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0063:score", "group_id": "wf12:utility_maintenance:test:0063", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 400ef8388a.\\nA means isolation >= 60.\\nD means emergency >= 75.\\nC means permit >= 85.\\nB means crew >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": 88, \\\"permit\\\": 93, \\\"emergency\\\": 47}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0063:noul", "group_id": "wf12:utility_maintenance:test:0063", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 400ef8388a.\\nA means isolation >= 60.\\nD means emergency >= 75.\\nC means permit >= 85.\\nB means crew >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": 88, \\\"permit\\\": 93, \\\"emergency\\\": 47}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0063:uncertain", "group_id": "wf12:utility_maintenance:test:0063", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 400ef8388a.\\nA means isolation >= 60.\\nD means emergency >= 75.\\nC means permit >= 85.\\nB means crew >= 41.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.05, \\\"C\\\": 0.95, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9073749999999999, \"true\": 0.09262500000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0064:choice", "group_id": "wf12:utility_maintenance:test:0064", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c0b208bd04.\\nC means permit >= 60.\\nA means isolation >= 20.\\nD means emergency >= 81.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 69, \\\"crew\\\": 3, \\\"permit\\\": 78, \\\"emergency\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9a03a1\": \"Outcome PROCEED.\", \"candidate_be0e30\": \"Outcome REVIEW.\", \"candidate_219f37\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9a03a1\": 0.0, \"candidate_be0e30\": 1.0, \"candidate_219f37\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0064:score", "group_id": "wf12:utility_maintenance:test:0064", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c0b208bd04.\\nC means permit >= 60.\\nA means isolation >= 20.\\nD means emergency >= 81.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 69, \\\"crew\\\": 3, \\\"permit\\\": 78, \\\"emergency\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0064:noul", "group_id": "wf12:utility_maintenance:test:0064", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c0b208bd04.\\nC means permit >= 60.\\nA means isolation >= 20.\\nD means emergency >= 81.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 69, \\\"crew\\\": 3, \\\"permit\\\": 78, \\\"emergency\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0064:uncertain", "group_id": "wf12:utility_maintenance:test:0064", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c0b208bd04.\\nC means permit >= 60.\\nA means isolation >= 20.\\nD means emergency >= 81.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 3, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 23}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.7725, \"true\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0065:choice", "group_id": "wf12:utility_maintenance:test:0065", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 04dc19b922.\\nD means emergency >= 25.\\nC means permit >= 46.\\nB means crew >= 52.\\nA means isolation >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 9, \\\"crew\\\": 17, \\\"permit\\\": 35, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_41e4d0\": \"Outcome PROCEED.\", \"candidate_fc506c\": \"Outcome REVIEW.\", \"candidate_5b2fc5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_41e4d0\": 0.0, \"candidate_fc506c\": 0.0, \"candidate_5b2fc5\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0065:score", "group_id": "wf12:utility_maintenance:test:0065", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 04dc19b922.\\nD means emergency >= 25.\\nC means permit >= 46.\\nB means crew >= 52.\\nA means isolation >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 9, \\\"crew\\\": 17, \\\"permit\\\": 35, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0065:noul", "group_id": "wf12:utility_maintenance:test:0065", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 04dc19b922.\\nD means emergency >= 25.\\nC means permit >= 46.\\nB means crew >= 52.\\nA means isolation >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 9, \\\"crew\\\": 17, \\\"permit\\\": 35, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0065:uncertain", "group_id": "wf12:utility_maintenance:test:0065", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 04dc19b922.\\nD means emergency >= 25.\\nC means permit >= 46.\\nB means crew >= 52.\\nA means isolation >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 35, \\\"emergency\\\": 7}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0066:choice", "group_id": "wf12:utility_maintenance:test:0066", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2f3f5e431d.\\nC means permit >= 89.\\nB means crew >= 61.\\nD means emergency >= 79.\\nA means isolation >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 39, \\\"crew\\\": 83, \\\"permit\\\": 94, \\\"emergency\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_09fe1d\": \"Outcome PROCEED.\", \"candidate_d76bcd\": \"Outcome REVIEW.\", \"candidate_a4a310\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_09fe1d\": 1.0, \"candidate_d76bcd\": 0.0, \"candidate_a4a310\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0066:score", "group_id": "wf12:utility_maintenance:test:0066", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2f3f5e431d.\\nC means permit >= 89.\\nB means crew >= 61.\\nD means emergency >= 79.\\nA means isolation >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 39, \\\"crew\\\": 83, \\\"permit\\\": 94, \\\"emergency\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0066:noul", "group_id": "wf12:utility_maintenance:test:0066", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2f3f5e431d.\\nC means permit >= 89.\\nB means crew >= 61.\\nD means emergency >= 79.\\nA means isolation >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 39, \\\"crew\\\": 83, \\\"permit\\\": 94, \\\"emergency\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0066:uncertain", "group_id": "wf12:utility_maintenance:test:0066", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2f3f5e431d.\\nC means permit >= 89.\\nB means crew >= 61.\\nD means emergency >= 79.\\nA means isolation >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 39, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 94, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_09fe1d\": \"Outcome PROCEED.\", \"candidate_d76bcd\": \"Outcome REVIEW.\", \"candidate_a4a310\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_09fe1d\": 0.995, \"candidate_d76bcd\": 0.005000000000000004, \"candidate_a4a310\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0067:choice", "group_id": "wf12:utility_maintenance:test:0067", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a3650a1713.\\nD means emergency >= 14.\\nC means permit >= 80.\\nA means isolation >= 73.\\nB means crew >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 36, \\\"crew\\\": 0, \\\"permit\\\": 92, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1f8900\": \"Outcome PROCEED.\", \"candidate_97323b\": \"Outcome REVIEW.\", \"candidate_e63294\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1f8900\": 0.0, \"candidate_97323b\": 1.0, \"candidate_e63294\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0067:score", "group_id": "wf12:utility_maintenance:test:0067", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a3650a1713.\\nD means emergency >= 14.\\nC means permit >= 80.\\nA means isolation >= 73.\\nB means crew >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 36, \\\"crew\\\": 0, \\\"permit\\\": 92, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0067:noul", "group_id": "wf12:utility_maintenance:test:0067", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a3650a1713.\\nD means emergency >= 14.\\nC means permit >= 80.\\nA means isolation >= 73.\\nB means crew >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 36, \\\"crew\\\": 0, \\\"permit\\\": 92, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0067:uncertain", "group_id": "wf12:utility_maintenance:test:0067", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a3650a1713.\\nD means emergency >= 14.\\nC means permit >= 80.\\nA means isolation >= 73.\\nB means crew >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 36, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 2}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"C\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1f8900\": \"Outcome PROCEED.\", \"candidate_97323b\": \"Outcome REVIEW.\", \"candidate_e63294\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1f8900\": 0.0, \"candidate_97323b\": 0.35, \"candidate_e63294\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0068:choice", "group_id": "wf12:utility_maintenance:test:0068", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 14fc47454c.\\nD means emergency >= 18.\\nB means crew >= 77.\\nC means permit >= 62.\\nA means isolation >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 12, \\\"crew\\\": 20, \\\"permit\\\": 64, \\\"emergency\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_13ede1\": \"Outcome PROCEED.\", \"candidate_b1c71f\": \"Outcome REVIEW.\", \"candidate_84aee0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_13ede1\": 0.0, \"candidate_b1c71f\": 0.0, \"candidate_84aee0\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0068:score", "group_id": "wf12:utility_maintenance:test:0068", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 14fc47454c.\\nD means emergency >= 18.\\nB means crew >= 77.\\nC means permit >= 62.\\nA means isolation >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 12, \\\"crew\\\": 20, \\\"permit\\\": 64, \\\"emergency\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0068:noul", "group_id": "wf12:utility_maintenance:test:0068", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 14fc47454c.\\nD means emergency >= 18.\\nB means crew >= 77.\\nC means permit >= 62.\\nA means isolation >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 12, \\\"crew\\\": 20, \\\"permit\\\": 64, \\\"emergency\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0068:uncertain", "group_id": "wf12:utility_maintenance:test:0068", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 14fc47454c.\\nD means emergency >= 18.\\nB means crew >= 77.\\nC means permit >= 62.\\nA means isolation >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 64, \\\"emergency\\\": 37}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_13ede1\": \"Outcome PROCEED.\", \"candidate_b1c71f\": \"Outcome REVIEW.\", \"candidate_84aee0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_13ede1\": 0.35, \"candidate_b1c71f\": 0.0, \"candidate_84aee0\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0069:choice", "group_id": "wf12:utility_maintenance:test:0069", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 018418e47a.\\nA means isolation >= 34.\\nC means permit >= 43.\\nB means crew >= 84.\\nD means emergency >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 92, \\\"permit\\\": 46, \\\"emergency\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_526fa3\": \"Outcome PROCEED.\", \"candidate_c0c305\": \"Outcome REVIEW.\", \"candidate_17eb57\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_526fa3\": 1.0, \"candidate_c0c305\": 0.0, \"candidate_17eb57\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0069:score", "group_id": "wf12:utility_maintenance:test:0069", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 018418e47a.\\nA means isolation >= 34.\\nC means permit >= 43.\\nB means crew >= 84.\\nD means emergency >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 92, \\\"permit\\\": 46, \\\"emergency\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0069:noul", "group_id": "wf12:utility_maintenance:test:0069", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 018418e47a.\\nA means isolation >= 34.\\nC means permit >= 43.\\nB means crew >= 84.\\nD means emergency >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 92, \\\"permit\\\": 46, \\\"emergency\\\": 11}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0069:uncertain", "group_id": "wf12:utility_maintenance:test:0069", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 018418e47a.\\nA means isolation >= 34.\\nC means permit >= 43.\\nB means crew >= 84.\\nD means emergency >= 33.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 46, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9525, \"1\": 0.04750000000000004, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0070:choice", "group_id": "wf12:utility_maintenance:test:0070", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 40d1f82efb.\\nC means permit >= 13.\\nA means isolation >= 79.\\nB means crew >= 71.\\nD means emergency >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 81, \\\"crew\\\": 37, \\\"permit\\\": 7, \\\"emergency\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_703391\": \"Outcome PROCEED.\", \"candidate_aca557\": \"Outcome REVIEW.\", \"candidate_b06178\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_703391\": 0.0, \"candidate_aca557\": 1.0, \"candidate_b06178\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0070:score", "group_id": "wf12:utility_maintenance:test:0070", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 40d1f82efb.\\nC means permit >= 13.\\nA means isolation >= 79.\\nB means crew >= 71.\\nD means emergency >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 81, \\\"crew\\\": 37, \\\"permit\\\": 7, \\\"emergency\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0070:noul", "group_id": "wf12:utility_maintenance:test:0070", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 40d1f82efb.\\nC means permit >= 13.\\nA means isolation >= 79.\\nB means crew >= 71.\\nD means emergency >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 81, \\\"crew\\\": 37, \\\"permit\\\": 7, \\\"emergency\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0070:uncertain", "group_id": "wf12:utility_maintenance:test:0070", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 40d1f82efb.\\nC means permit >= 13.\\nA means isolation >= 79.\\nB means crew >= 71.\\nD means emergency >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 81, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 7, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.65, \"1\": 0.35, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0071:choice", "group_id": "wf12:utility_maintenance:test:0071", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2619bbb4b3.\\nD means emergency >= 82.\\nB means crew >= 13.\\nA means isolation >= 52.\\nC means permit >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 15, \\\"crew\\\": 82, \\\"permit\\\": 30, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4b8a26\": \"Outcome PROCEED.\", \"candidate_b65bef\": \"Outcome REVIEW.\", \"candidate_4c485e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4b8a26\": 0.0, \"candidate_b65bef\": 0.0, \"candidate_4c485e\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0071:score", "group_id": "wf12:utility_maintenance:test:0071", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2619bbb4b3.\\nD means emergency >= 82.\\nB means crew >= 13.\\nA means isolation >= 52.\\nC means permit >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 15, \\\"crew\\\": 82, \\\"permit\\\": 30, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0071:noul", "group_id": "wf12:utility_maintenance:test:0071", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2619bbb4b3.\\nD means emergency >= 82.\\nB means crew >= 13.\\nA means isolation >= 52.\\nC means permit >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 15, \\\"crew\\\": 82, \\\"permit\\\": 30, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0071:uncertain", "group_id": "wf12:utility_maintenance:test:0071", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2619bbb4b3.\\nD means emergency >= 82.\\nB means crew >= 13.\\nA means isolation >= 52.\\nC means permit >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 82, \\\"permit\\\": 30, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.8960000000000001, \"1\": 0.0831999999999999, \"2\": 0.020799999999999975}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0072:choice", "group_id": "wf12:utility_maintenance:test:0072", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 62c74061dd.\\nD means emergency >= 30.\\nB means crew >= 70.\\nA means isolation >= 69.\\nC means permit >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 73, \\\"crew\\\": 84, \\\"permit\\\": 4, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d67e12\": \"Outcome PROCEED.\", \"candidate_78e931\": \"Outcome REVIEW.\", \"candidate_738c71\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d67e12\": 1.0, \"candidate_78e931\": 0.0, \"candidate_738c71\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0072:score", "group_id": "wf12:utility_maintenance:test:0072", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 62c74061dd.\\nD means emergency >= 30.\\nB means crew >= 70.\\nA means isolation >= 69.\\nC means permit >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 73, \\\"crew\\\": 84, \\\"permit\\\": 4, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0072:noul", "group_id": "wf12:utility_maintenance:test:0072", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 62c74061dd.\\nD means emergency >= 30.\\nB means crew >= 70.\\nA means isolation >= 69.\\nC means permit >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 73, \\\"crew\\\": 84, \\\"permit\\\": 4, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0072:uncertain", "group_id": "wf12:utility_maintenance:test:0072", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 62c74061dd.\\nD means emergency >= 30.\\nB means crew >= 70.\\nA means isolation >= 69.\\nC means permit >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.95, \\\"C\\\": 0.05, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.10450000000000004, \"true\": 0.8955}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0073:choice", "group_id": "wf12:utility_maintenance:test:0073", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eb1ae66d44.\\nA means isolation >= 86.\\nD means emergency >= 27.\\nB means crew >= 62.\\nC means permit >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 0, \\\"permit\\\": 32, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_130beb\": \"Outcome PROCEED.\", \"candidate_ade3fd\": \"Outcome REVIEW.\", \"candidate_557edf\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_130beb\": 0.0, \"candidate_ade3fd\": 1.0, \"candidate_557edf\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0073:score", "group_id": "wf12:utility_maintenance:test:0073", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eb1ae66d44.\\nA means isolation >= 86.\\nD means emergency >= 27.\\nB means crew >= 62.\\nC means permit >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 0, \\\"permit\\\": 32, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0073:noul", "group_id": "wf12:utility_maintenance:test:0073", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eb1ae66d44.\\nA means isolation >= 86.\\nD means emergency >= 27.\\nB means crew >= 62.\\nC means permit >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 0, \\\"permit\\\": 32, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0073:uncertain", "group_id": "wf12:utility_maintenance:test:0073", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eb1ae66d44.\\nA means isolation >= 86.\\nD means emergency >= 27.\\nB means crew >= 62.\\nC means permit >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.65, \\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8455, \"true\": 0.15449999999999997}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0074:choice", "group_id": "wf12:utility_maintenance:test:0074", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ca6715115d.\\nC means permit >= 56.\\nB means crew >= 22.\\nA means isolation >= 81.\\nD means emergency >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 26, \\\"crew\\\": 3, \\\"permit\\\": 77, \\\"emergency\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fe5eb7\": \"Outcome PROCEED.\", \"candidate_3321cd\": \"Outcome REVIEW.\", \"candidate_22bb7a\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fe5eb7\": 0.0, \"candidate_3321cd\": 0.0, \"candidate_22bb7a\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0074:score", "group_id": "wf12:utility_maintenance:test:0074", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ca6715115d.\\nC means permit >= 56.\\nB means crew >= 22.\\nA means isolation >= 81.\\nD means emergency >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 26, \\\"crew\\\": 3, \\\"permit\\\": 77, \\\"emergency\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0074:noul", "group_id": "wf12:utility_maintenance:test:0074", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ca6715115d.\\nC means permit >= 56.\\nB means crew >= 22.\\nA means isolation >= 81.\\nD means emergency >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 26, \\\"crew\\\": 3, \\\"permit\\\": 77, \\\"emergency\\\": 89}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0074:uncertain", "group_id": "wf12:utility_maintenance:test:0074", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ca6715115d.\\nC means permit >= 56.\\nB means crew >= 22.\\nA means isolation >= 81.\\nD means emergency >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 3, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 89}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8960000000000001, \"true\": 0.10399999999999987}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0075:choice", "group_id": "wf12:utility_maintenance:test:0075", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1f21d6572c.\\nD means emergency >= 67.\\nC means permit >= 55.\\nA means isolation >= 58.\\nB means crew >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": 83, \\\"permit\\\": 54, \\\"emergency\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_37a5d4\": \"Outcome PROCEED.\", \"candidate_3c5df0\": \"Outcome REVIEW.\", \"candidate_9a07a0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_37a5d4\": 1.0, \"candidate_3c5df0\": 0.0, \"candidate_9a07a0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0075:score", "group_id": "wf12:utility_maintenance:test:0075", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1f21d6572c.\\nD means emergency >= 67.\\nC means permit >= 55.\\nA means isolation >= 58.\\nB means crew >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": 83, \\\"permit\\\": 54, \\\"emergency\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0075:noul", "group_id": "wf12:utility_maintenance:test:0075", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1f21d6572c.\\nD means emergency >= 67.\\nC means permit >= 55.\\nA means isolation >= 58.\\nB means crew >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": 83, \\\"permit\\\": 54, \\\"emergency\\\": 76}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0075:uncertain", "group_id": "wf12:utility_maintenance:test:0075", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 1f21d6572c.\\nD means emergency >= 67.\\nC means permit >= 55.\\nA means isolation >= 58.\\nB means crew >= 46.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 54, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.9, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_37a5d4\": \"Outcome PROCEED.\", \"candidate_3c5df0\": \"Outcome REVIEW.\", \"candidate_9a07a0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_37a5d4\": 0.9, \"candidate_3c5df0\": 0.09999999999999998, \"candidate_9a07a0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0076:choice", "group_id": "wf12:utility_maintenance:test:0076", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eb9210fe95.\\nA means isolation >= 10.\\nB means crew >= 68.\\nD means emergency >= 88.\\nC means permit >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 33, \\\"crew\\\": 62, \\\"permit\\\": 42, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2236e8\": \"Outcome PROCEED.\", \"candidate_b1403c\": \"Outcome REVIEW.\", \"candidate_739983\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2236e8\": 0.0, \"candidate_b1403c\": 1.0, \"candidate_739983\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0076:score", "group_id": "wf12:utility_maintenance:test:0076", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eb9210fe95.\\nA means isolation >= 10.\\nB means crew >= 68.\\nD means emergency >= 88.\\nC means permit >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 33, \\\"crew\\\": 62, \\\"permit\\\": 42, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0076:noul", "group_id": "wf12:utility_maintenance:test:0076", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eb9210fe95.\\nA means isolation >= 10.\\nB means crew >= 68.\\nD means emergency >= 88.\\nC means permit >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 33, \\\"crew\\\": 62, \\\"permit\\\": 42, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0076:uncertain", "group_id": "wf12:utility_maintenance:test:0076", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eb9210fe95.\\nA means isolation >= 10.\\nB means crew >= 68.\\nD means emergency >= 88.\\nC means permit >= 44.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 62, \\\"permit\\\": 42, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2236e8\": \"Outcome PROCEED.\", \"candidate_b1403c\": \"Outcome REVIEW.\", \"candidate_739983\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2236e8\": 0.0, \"candidate_b1403c\": 0.35, \"candidate_739983\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0077:choice", "group_id": "wf12:utility_maintenance:test:0077", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3d5020f106.\\nC means permit >= 80.\\nA means isolation >= 88.\\nB means crew >= 24.\\nD means emergency >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 17, \\\"crew\\\": 53, \\\"permit\\\": 17, \\\"emergency\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f80b6e\": \"Outcome PROCEED.\", \"candidate_e6c425\": \"Outcome REVIEW.\", \"candidate_4ccfe4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f80b6e\": 0.0, \"candidate_e6c425\": 0.0, \"candidate_4ccfe4\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0077:score", "group_id": "wf12:utility_maintenance:test:0077", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3d5020f106.\\nC means permit >= 80.\\nA means isolation >= 88.\\nB means crew >= 24.\\nD means emergency >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 17, \\\"crew\\\": 53, \\\"permit\\\": 17, \\\"emergency\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0077:noul", "group_id": "wf12:utility_maintenance:test:0077", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3d5020f106.\\nC means permit >= 80.\\nA means isolation >= 88.\\nB means crew >= 24.\\nD means emergency >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 17, \\\"crew\\\": 53, \\\"permit\\\": 17, \\\"emergency\\\": 35}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0077:uncertain", "group_id": "wf12:utility_maintenance:test:0077", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3d5020f106.\\nC means permit >= 80.\\nA means isolation >= 88.\\nB means crew >= 24.\\nD means emergency >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 53, \\\"permit\\\": 17, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f80b6e\": \"Outcome PROCEED.\", \"candidate_e6c425\": \"Outcome REVIEW.\", \"candidate_4ccfe4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f80b6e\": 0.35, \"candidate_e6c425\": 0.0, \"candidate_4ccfe4\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0078:choice", "group_id": "wf12:utility_maintenance:test:0078", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4f5ae55e84.\\nB means crew >= 86.\\nA means isolation >= 37.\\nD means emergency >= 29.\\nC means permit >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 42, \\\"crew\\\": 100, \\\"permit\\\": 80, \\\"emergency\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_119811\": \"Outcome PROCEED.\", \"candidate_1cf563\": \"Outcome REVIEW.\", \"candidate_4d41a3\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_119811\": 1.0, \"candidate_1cf563\": 0.0, \"candidate_4d41a3\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0078:score", "group_id": "wf12:utility_maintenance:test:0078", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4f5ae55e84.\\nB means crew >= 86.\\nA means isolation >= 37.\\nD means emergency >= 29.\\nC means permit >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 42, \\\"crew\\\": 100, \\\"permit\\\": 80, \\\"emergency\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0078:noul", "group_id": "wf12:utility_maintenance:test:0078", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4f5ae55e84.\\nB means crew >= 86.\\nA means isolation >= 37.\\nD means emergency >= 29.\\nC means permit >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 42, \\\"crew\\\": 100, \\\"permit\\\": 80, \\\"emergency\\\": 13}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0078:uncertain", "group_id": "wf12:utility_maintenance:test:0078", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4f5ae55e84.\\nB means crew >= 86.\\nA means isolation >= 37.\\nD means emergency >= 29.\\nC means permit >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 80, \\\"emergency\\\": 13}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.005000000000000001, \"1\": 0.995, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0079:choice", "group_id": "wf12:utility_maintenance:test:0079", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3aa5d521b5.\\nB means crew >= 16.\\nA means isolation >= 26.\\nC means permit >= 72.\\nD means emergency >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 16, \\\"crew\\\": 83, \\\"permit\\\": 87, \\\"emergency\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_37d2d8\": \"Outcome PROCEED.\", \"candidate_08155d\": \"Outcome REVIEW.\", \"candidate_11ff0d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_37d2d8\": 0.0, \"candidate_08155d\": 1.0, \"candidate_11ff0d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0079:score", "group_id": "wf12:utility_maintenance:test:0079", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3aa5d521b5.\\nB means crew >= 16.\\nA means isolation >= 26.\\nC means permit >= 72.\\nD means emergency >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 16, \\\"crew\\\": 83, \\\"permit\\\": 87, \\\"emergency\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0079:noul", "group_id": "wf12:utility_maintenance:test:0079", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3aa5d521b5.\\nB means crew >= 16.\\nA means isolation >= 26.\\nC means permit >= 72.\\nD means emergency >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 16, \\\"crew\\\": 83, \\\"permit\\\": 87, \\\"emergency\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0079:uncertain", "group_id": "wf12:utility_maintenance:test:0079", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3aa5d521b5.\\nB means crew >= 16.\\nA means isolation >= 26.\\nC means permit >= 72.\\nD means emergency >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 83, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 42}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.42250000000000004, \"2\": 0.22749999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0080:choice", "group_id": "wf12:utility_maintenance:test:0080", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference b1973cbe7c.\\nA means isolation >= 90.\\nC means permit >= 53.\\nD means emergency >= 65.\\nB means crew >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 90, \\\"permit\\\": 60, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_676d82\": \"Outcome PROCEED.\", \"candidate_d92058\": \"Outcome REVIEW.\", \"candidate_940111\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_676d82\": 0.0, \"candidate_d92058\": 0.0, \"candidate_940111\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0080:score", "group_id": "wf12:utility_maintenance:test:0080", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference b1973cbe7c.\\nA means isolation >= 90.\\nC means permit >= 53.\\nD means emergency >= 65.\\nB means crew >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 90, \\\"permit\\\": 60, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0080:noul", "group_id": "wf12:utility_maintenance:test:0080", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference b1973cbe7c.\\nA means isolation >= 90.\\nC means permit >= 53.\\nD means emergency >= 65.\\nB means crew >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 90, \\\"permit\\\": 60, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0080:uncertain", "group_id": "wf12:utility_maintenance:test:0080", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference b1973cbe7c.\\nA means isolation >= 90.\\nC means permit >= 53.\\nD means emergency >= 65.\\nB means crew >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 60, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.19999999999999996, \"2\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0081:choice", "group_id": "wf12:utility_maintenance:test:0081", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eeb74fcf2c.\\nC means permit >= 15.\\nA means isolation >= 19.\\nD means emergency >= 89.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 46, \\\"permit\\\": 7, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1e386e\": \"Outcome PROCEED.\", \"candidate_a090f7\": \"Outcome REVIEW.\", \"candidate_fdcf5e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1e386e\": 1.0, \"candidate_a090f7\": 0.0, \"candidate_fdcf5e\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0081:score", "group_id": "wf12:utility_maintenance:test:0081", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eeb74fcf2c.\\nC means permit >= 15.\\nA means isolation >= 19.\\nD means emergency >= 89.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 46, \\\"permit\\\": 7, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0081:noul", "group_id": "wf12:utility_maintenance:test:0081", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eeb74fcf2c.\\nC means permit >= 15.\\nA means isolation >= 19.\\nD means emergency >= 89.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 46, \\\"permit\\\": 7, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0081:uncertain", "group_id": "wf12:utility_maintenance:test:0081", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eeb74fcf2c.\\nC means permit >= 15.\\nA means isolation >= 19.\\nD means emergency >= 89.\\nB means crew >= 38.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.9, \\\"B\\\": 0.9, \\\"C\\\": 0.05, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9005000000000001, \"true\": 0.09949999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0082:choice", "group_id": "wf12:utility_maintenance:test:0082", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31df03d004.\\nC means permit >= 26.\\nA means isolation >= 74.\\nB means crew >= 48.\\nD means emergency >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 22, \\\"crew\\\": 42, \\\"permit\\\": 54, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_200f79\": \"Outcome PROCEED.\", \"candidate_fddc67\": \"Outcome REVIEW.\", \"candidate_2169d8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_200f79\": 0.0, \"candidate_fddc67\": 1.0, \"candidate_2169d8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0082:score", "group_id": "wf12:utility_maintenance:test:0082", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31df03d004.\\nC means permit >= 26.\\nA means isolation >= 74.\\nB means crew >= 48.\\nD means emergency >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 22, \\\"crew\\\": 42, \\\"permit\\\": 54, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0082:noul", "group_id": "wf12:utility_maintenance:test:0082", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31df03d004.\\nC means permit >= 26.\\nA means isolation >= 74.\\nB means crew >= 48.\\nD means emergency >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 22, \\\"crew\\\": 42, \\\"permit\\\": 54, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0082:uncertain", "group_id": "wf12:utility_maintenance:test:0082", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 31df03d004.\\nC means permit >= 26.\\nA means isolation >= 74.\\nB means crew >= 48.\\nD means emergency >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 42, \\\"permit\\\": 54, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.87, \"true\": 0.13}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0083:choice", "group_id": "wf12:utility_maintenance:test:0083", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c77ea5d96b.\\nA means isolation >= 32.\\nC means permit >= 36.\\nD means emergency >= 18.\\nB means crew >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 8, \\\"crew\\\": 74, \\\"permit\\\": 46, \\\"emergency\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a7446\": \"Outcome PROCEED.\", \"candidate_bfb8c7\": \"Outcome REVIEW.\", \"candidate_dbd7d5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a7446\": 0.0, \"candidate_bfb8c7\": 0.0, \"candidate_dbd7d5\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0083:score", "group_id": "wf12:utility_maintenance:test:0083", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c77ea5d96b.\\nA means isolation >= 32.\\nC means permit >= 36.\\nD means emergency >= 18.\\nB means crew >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 8, \\\"crew\\\": 74, \\\"permit\\\": 46, \\\"emergency\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0083:noul", "group_id": "wf12:utility_maintenance:test:0083", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c77ea5d96b.\\nA means isolation >= 32.\\nC means permit >= 36.\\nD means emergency >= 18.\\nB means crew >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 8, \\\"crew\\\": 74, \\\"permit\\\": 46, \\\"emergency\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0083:uncertain", "group_id": "wf12:utility_maintenance:test:0083", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c77ea5d96b.\\nA means isolation >= 32.\\nC means permit >= 36.\\nD means emergency >= 18.\\nB means crew >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.2, \\\"C\\\": 0.2, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.6745, \"true\": 0.3255}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0084:choice", "group_id": "wf12:utility_maintenance:test:0084", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd0bf462c9.\\nA means isolation >= 53.\\nD means emergency >= 15.\\nC means permit >= 72.\\nB means crew >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 76, \\\"crew\\\": 86, \\\"permit\\\": 71, \\\"emergency\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_287f28\": \"Outcome PROCEED.\", \"candidate_bb7814\": \"Outcome REVIEW.\", \"candidate_f30d8e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_287f28\": 1.0, \"candidate_bb7814\": 0.0, \"candidate_f30d8e\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0084:score", "group_id": "wf12:utility_maintenance:test:0084", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd0bf462c9.\\nA means isolation >= 53.\\nD means emergency >= 15.\\nC means permit >= 72.\\nB means crew >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 76, \\\"crew\\\": 86, \\\"permit\\\": 71, \\\"emergency\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0084:noul", "group_id": "wf12:utility_maintenance:test:0084", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd0bf462c9.\\nA means isolation >= 53.\\nD means emergency >= 15.\\nC means permit >= 72.\\nB means crew >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 76, \\\"crew\\\": 86, \\\"permit\\\": 71, \\\"emergency\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0084:uncertain", "group_id": "wf12:utility_maintenance:test:0084", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd0bf462c9.\\nA means isolation >= 53.\\nD means emergency >= 15.\\nC means permit >= 72.\\nB means crew >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 71, \\\"emergency\\\": 18}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_287f28\": \"Outcome PROCEED.\", \"candidate_bb7814\": \"Outcome REVIEW.\", \"candidate_f30d8e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_287f28\": 0.005000000000000001, \"candidate_bb7814\": 0.095, \"candidate_f30d8e\": 0.9}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0085:choice", "group_id": "wf12:utility_maintenance:test:0085", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 00cb761361.\\nA means isolation >= 75.\\nC means permit >= 69.\\nB means crew >= 66.\\nD means emergency >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 44, \\\"permit\\\": 39, \\\"emergency\\\": 65}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bab8e3\": \"Outcome PROCEED.\", \"candidate_32a023\": \"Outcome REVIEW.\", \"candidate_0c281d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bab8e3\": 0.0, \"candidate_32a023\": 1.0, \"candidate_0c281d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0085:score", "group_id": "wf12:utility_maintenance:test:0085", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 00cb761361.\\nA means isolation >= 75.\\nC means permit >= 69.\\nB means crew >= 66.\\nD means emergency >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 44, \\\"permit\\\": 39, \\\"emergency\\\": 65}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0085:noul", "group_id": "wf12:utility_maintenance:test:0085", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 00cb761361.\\nA means isolation >= 75.\\nC means permit >= 69.\\nB means crew >= 66.\\nD means emergency >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 44, \\\"permit\\\": 39, \\\"emergency\\\": 65}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0085:uncertain", "group_id": "wf12:utility_maintenance:test:0085", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 00cb761361.\\nA means isolation >= 75.\\nC means permit >= 69.\\nB means crew >= 66.\\nD means emergency >= 84.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 91, \\\"crew\\\": 44, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bab8e3\": \"Outcome PROCEED.\", \"candidate_32a023\": \"Outcome REVIEW.\", \"candidate_0c281d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bab8e3\": 0.22749999999999998, \"candidate_32a023\": 0.7725, \"candidate_0c281d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0086:choice", "group_id": "wf12:utility_maintenance:test:0086", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 84a3d72503.\\nC means permit >= 55.\\nB means crew >= 52.\\nD means emergency >= 74.\\nA means isolation >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 21, \\\"crew\\\": 8, \\\"permit\\\": 94, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a8b09f\": \"Outcome PROCEED.\", \"candidate_b4eb91\": \"Outcome REVIEW.\", \"candidate_2edba5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a8b09f\": 0.0, \"candidate_b4eb91\": 0.0, \"candidate_2edba5\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0086:score", "group_id": "wf12:utility_maintenance:test:0086", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 84a3d72503.\\nC means permit >= 55.\\nB means crew >= 52.\\nD means emergency >= 74.\\nA means isolation >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 21, \\\"crew\\\": 8, \\\"permit\\\": 94, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0086:noul", "group_id": "wf12:utility_maintenance:test:0086", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 84a3d72503.\\nC means permit >= 55.\\nB means crew >= 52.\\nD means emergency >= 74.\\nA means isolation >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 21, \\\"crew\\\": 8, \\\"permit\\\": 94, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0086:uncertain", "group_id": "wf12:utility_maintenance:test:0086", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 84a3d72503.\\nC means permit >= 55.\\nB means crew >= 52.\\nD means emergency >= 74.\\nA means isolation >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 8, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 98}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a8b09f\": \"Outcome PROCEED.\", \"candidate_b4eb91\": \"Outcome REVIEW.\", \"candidate_2edba5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a8b09f\": 0.10739726027397259, \"candidate_b4eb91\": 0.42958904109589036, \"candidate_2edba5\": 0.4630136986301371}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0087:choice", "group_id": "wf12:utility_maintenance:test:0087", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 25e7e67d53.\\nA means isolation >= 87.\\nD means emergency >= 86.\\nC means permit >= 29.\\nB means crew >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 90, \\\"permit\\\": 28, \\\"emergency\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f9d7f2\": \"Outcome PROCEED.\", \"candidate_195cb6\": \"Outcome REVIEW.\", \"candidate_56308f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f9d7f2\": 1.0, \"candidate_195cb6\": 0.0, \"candidate_56308f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0087:score", "group_id": "wf12:utility_maintenance:test:0087", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 25e7e67d53.\\nA means isolation >= 87.\\nD means emergency >= 86.\\nC means permit >= 29.\\nB means crew >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 90, \\\"permit\\\": 28, \\\"emergency\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0087:noul", "group_id": "wf12:utility_maintenance:test:0087", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 25e7e67d53.\\nA means isolation >= 87.\\nD means emergency >= 86.\\nC means permit >= 29.\\nB means crew >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 90, \\\"permit\\\": 28, \\\"emergency\\\": 23}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0087:uncertain", "group_id": "wf12:utility_maintenance:test:0087", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 25e7e67d53.\\nA means isolation >= 87.\\nD means emergency >= 86.\\nC means permit >= 29.\\nB means crew >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 23}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.95, \"1\": 0.050000000000000044, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0088:choice", "group_id": "wf12:utility_maintenance:test:0088", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7caa93bf2a.\\nC means permit >= 39.\\nD means emergency >= 89.\\nA means isolation >= 88.\\nB means crew >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 0, \\\"crew\\\": 89, \\\"permit\\\": 76, \\\"emergency\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cb23cb\": \"Outcome PROCEED.\", \"candidate_ed8b91\": \"Outcome REVIEW.\", \"candidate_156b0c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cb23cb\": 0.0, \"candidate_ed8b91\": 1.0, \"candidate_156b0c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0088:score", "group_id": "wf12:utility_maintenance:test:0088", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7caa93bf2a.\\nC means permit >= 39.\\nD means emergency >= 89.\\nA means isolation >= 88.\\nB means crew >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 0, \\\"crew\\\": 89, \\\"permit\\\": 76, \\\"emergency\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0088:noul", "group_id": "wf12:utility_maintenance:test:0088", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7caa93bf2a.\\nC means permit >= 39.\\nD means emergency >= 89.\\nA means isolation >= 88.\\nB means crew >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 0, \\\"crew\\\": 89, \\\"permit\\\": 76, \\\"emergency\\\": 51}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0088:uncertain", "group_id": "wf12:utility_maintenance:test:0088", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 7caa93bf2a.\\nC means permit >= 39.\\nD means emergency >= 89.\\nA means isolation >= 88.\\nB means crew >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 0, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 76, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.8, \"2\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0089:choice", "group_id": "wf12:utility_maintenance:test:0089", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4d472645b3.\\nC means permit >= 22.\\nA means isolation >= 40.\\nB means crew >= 56.\\nD means emergency >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 37, \\\"crew\\\": 12, \\\"permit\\\": 13, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f4e7bb\": \"Outcome PROCEED.\", \"candidate_ecafea\": \"Outcome REVIEW.\", \"candidate_8202c7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f4e7bb\": 0.0, \"candidate_ecafea\": 0.0, \"candidate_8202c7\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0089:score", "group_id": "wf12:utility_maintenance:test:0089", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4d472645b3.\\nC means permit >= 22.\\nA means isolation >= 40.\\nB means crew >= 56.\\nD means emergency >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 37, \\\"crew\\\": 12, \\\"permit\\\": 13, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0089:noul", "group_id": "wf12:utility_maintenance:test:0089", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4d472645b3.\\nC means permit >= 22.\\nA means isolation >= 40.\\nB means crew >= 56.\\nD means emergency >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 37, \\\"crew\\\": 12, \\\"permit\\\": 13, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0089:uncertain", "group_id": "wf12:utility_maintenance:test:0089", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 4d472645b3.\\nC means permit >= 22.\\nA means isolation >= 40.\\nB means crew >= 56.\\nD means emergency >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 12, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 6}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.9600000000000001, \"2\": 0.03999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0090:choice", "group_id": "wf12:utility_maintenance:test:0090", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e164d94d44.\\nC means permit >= 45.\\nB means crew >= 79.\\nA means isolation >= 34.\\nD means emergency >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 89, \\\"crew\\\": 51, \\\"permit\\\": 55, \\\"emergency\\\": 86}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e7105f\": \"Outcome PROCEED.\", \"candidate_24ced1\": \"Outcome REVIEW.\", \"candidate_b51114\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e7105f\": 1.0, \"candidate_24ced1\": 0.0, \"candidate_b51114\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0090:score", "group_id": "wf12:utility_maintenance:test:0090", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e164d94d44.\\nC means permit >= 45.\\nB means crew >= 79.\\nA means isolation >= 34.\\nD means emergency >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 89, \\\"crew\\\": 51, \\\"permit\\\": 55, \\\"emergency\\\": 86}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0090:noul", "group_id": "wf12:utility_maintenance:test:0090", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e164d94d44.\\nC means permit >= 45.\\nB means crew >= 79.\\nA means isolation >= 34.\\nD means emergency >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 89, \\\"crew\\\": 51, \\\"permit\\\": 55, \\\"emergency\\\": 86}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0090:uncertain", "group_id": "wf12:utility_maintenance:test:0090", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e164d94d44.\\nC means permit >= 45.\\nB means crew >= 79.\\nA means isolation >= 34.\\nD means emergency >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.05, \\\"C\\\": 0.05, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.14049999999999996, \"true\": 0.8595}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0091:choice", "group_id": "wf12:utility_maintenance:test:0091", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9c7f055f15.\\nD means emergency >= 63.\\nA means isolation >= 73.\\nB means crew >= 57.\\nC means permit >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 57, \\\"crew\\\": 89, \\\"permit\\\": 76, \\\"emergency\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ae1359\": \"Outcome PROCEED.\", \"candidate_d79aa7\": \"Outcome REVIEW.\", \"candidate_83973e\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ae1359\": 0.0, \"candidate_d79aa7\": 1.0, \"candidate_83973e\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0091:score", "group_id": "wf12:utility_maintenance:test:0091", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9c7f055f15.\\nD means emergency >= 63.\\nA means isolation >= 73.\\nB means crew >= 57.\\nC means permit >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 57, \\\"crew\\\": 89, \\\"permit\\\": 76, \\\"emergency\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0091:noul", "group_id": "wf12:utility_maintenance:test:0091", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9c7f055f15.\\nD means emergency >= 63.\\nA means isolation >= 73.\\nB means crew >= 57.\\nC means permit >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 57, \\\"crew\\\": 89, \\\"permit\\\": 76, \\\"emergency\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0091:uncertain", "group_id": "wf12:utility_maintenance:test:0091", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 9c7f055f15.\\nD means emergency >= 63.\\nA means isolation >= 73.\\nB means crew >= 57.\\nC means permit >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 89, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 4}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"C\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.3599999999999999, \"true\": 0.6400000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0092:choice", "group_id": "wf12:utility_maintenance:test:0092", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 8a906da321.\\nA means isolation >= 16.\\nC means permit >= 31.\\nB means crew >= 72.\\nD means emergency >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 0, \\\"crew\\\": 80, \\\"permit\\\": 2, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_50853a\": \"Outcome PROCEED.\", \"candidate_7edb66\": \"Outcome REVIEW.\", \"candidate_b31b13\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_50853a\": 0.0, \"candidate_7edb66\": 0.0, \"candidate_b31b13\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0092:score", "group_id": "wf12:utility_maintenance:test:0092", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 8a906da321.\\nA means isolation >= 16.\\nC means permit >= 31.\\nB means crew >= 72.\\nD means emergency >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 0, \\\"crew\\\": 80, \\\"permit\\\": 2, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0092:noul", "group_id": "wf12:utility_maintenance:test:0092", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 8a906da321.\\nA means isolation >= 16.\\nC means permit >= 31.\\nB means crew >= 72.\\nD means emergency >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 0, \\\"crew\\\": 80, \\\"permit\\\": 2, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0092:uncertain", "group_id": "wf12:utility_maintenance:test:0092", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 8a906da321.\\nA means isolation >= 16.\\nC means permit >= 31.\\nB means crew >= 72.\\nD means emergency >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.8, \\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.748, \"true\": 0.25200000000000006}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0093:choice", "group_id": "wf12:utility_maintenance:test:0093", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 486da61b0a.\\nD means emergency >= 53.\\nA means isolation >= 23.\\nC means permit >= 26.\\nB means crew >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 52, \\\"crew\\\": 94, \\\"permit\\\": 6, \\\"emergency\\\": 53}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_191d69\": \"Outcome PROCEED.\", \"candidate_7d97df\": \"Outcome REVIEW.\", \"candidate_6512bc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_191d69\": 1.0, \"candidate_7d97df\": 0.0, \"candidate_6512bc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0093:score", "group_id": "wf12:utility_maintenance:test:0093", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 486da61b0a.\\nD means emergency >= 53.\\nA means isolation >= 23.\\nC means permit >= 26.\\nB means crew >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 52, \\\"crew\\\": 94, \\\"permit\\\": 6, \\\"emergency\\\": 53}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0093:noul", "group_id": "wf12:utility_maintenance:test:0093", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 486da61b0a.\\nD means emergency >= 53.\\nA means isolation >= 23.\\nC means permit >= 26.\\nB means crew >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 52, \\\"crew\\\": 94, \\\"permit\\\": 6, \\\"emergency\\\": 53}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0093:uncertain", "group_id": "wf12:utility_maintenance:test:0093", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 486da61b0a.\\nD means emergency >= 53.\\nA means isolation >= 23.\\nC means permit >= 26.\\nB means crew >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 52, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 6, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_191d69\": \"Outcome PROCEED.\", \"candidate_7d97df\": \"Outcome REVIEW.\", \"candidate_6512bc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_191d69\": 0.05, \"candidate_7d97df\": 0.95, \"candidate_6512bc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0094:choice", "group_id": "wf12:utility_maintenance:test:0094", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ec9274a69b.\\nD means emergency >= 79.\\nA means isolation >= 14.\\nB means crew >= 41.\\nC means permit >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 19, \\\"permit\\\": 95, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2fa2c3\": \"Outcome PROCEED.\", \"candidate_f21314\": \"Outcome REVIEW.\", \"candidate_243adf\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2fa2c3\": 0.0, \"candidate_f21314\": 1.0, \"candidate_243adf\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0094:score", "group_id": "wf12:utility_maintenance:test:0094", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ec9274a69b.\\nD means emergency >= 79.\\nA means isolation >= 14.\\nB means crew >= 41.\\nC means permit >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 19, \\\"permit\\\": 95, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0094:noul", "group_id": "wf12:utility_maintenance:test:0094", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ec9274a69b.\\nD means emergency >= 79.\\nA means isolation >= 14.\\nB means crew >= 41.\\nC means permit >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 19, \\\"permit\\\": 95, \\\"emergency\\\": 20}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0094:uncertain", "group_id": "wf12:utility_maintenance:test:0094", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ec9274a69b.\\nD means emergency >= 79.\\nA means isolation >= 14.\\nB means crew >= 41.\\nC means permit >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 95, \\\"emergency\\\": 20}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2fa2c3\": \"Outcome PROCEED.\", \"candidate_f21314\": \"Outcome REVIEW.\", \"candidate_243adf\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2fa2c3\": 0.06999999999999999, \"candidate_f21314\": 0.93, \"candidate_243adf\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0095:choice", "group_id": "wf12:utility_maintenance:test:0095", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 09d130b756.\\nD means emergency >= 13.\\nB means crew >= 54.\\nC means permit >= 79.\\nA means isolation >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 21, \\\"crew\\\": 77, \\\"permit\\\": 36, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_272064\": \"Outcome PROCEED.\", \"candidate_cf67ce\": \"Outcome REVIEW.\", \"candidate_eaaadc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_272064\": 0.0, \"candidate_cf67ce\": 0.0, \"candidate_eaaadc\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0095:score", "group_id": "wf12:utility_maintenance:test:0095", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 09d130b756.\\nD means emergency >= 13.\\nB means crew >= 54.\\nC means permit >= 79.\\nA means isolation >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 21, \\\"crew\\\": 77, \\\"permit\\\": 36, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0095:noul", "group_id": "wf12:utility_maintenance:test:0095", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 09d130b756.\\nD means emergency >= 13.\\nB means crew >= 54.\\nC means permit >= 79.\\nA means isolation >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 21, \\\"crew\\\": 77, \\\"permit\\\": 36, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0095:uncertain", "group_id": "wf12:utility_maintenance:test:0095", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 09d130b756.\\nD means emergency >= 13.\\nB means crew >= 54.\\nC means permit >= 79.\\nA means isolation >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 21, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 2}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_272064\": \"Outcome PROCEED.\", \"candidate_cf67ce\": \"Outcome REVIEW.\", \"candidate_eaaadc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_272064\": 0.0, \"candidate_cf67ce\": 0.2, \"candidate_eaaadc\": 0.8}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0096:choice", "group_id": "wf12:utility_maintenance:test:0096", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c4812edd45.\\nA means isolation >= 61.\\nD means emergency >= 10.\\nC means permit >= 61.\\nB means crew >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": 81, \\\"permit\\\": 67, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7743ba\": \"Outcome PROCEED.\", \"candidate_f85d30\": \"Outcome REVIEW.\", \"candidate_38f24f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7743ba\": 1.0, \"candidate_f85d30\": 0.0, \"candidate_38f24f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0096:score", "group_id": "wf12:utility_maintenance:test:0096", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c4812edd45.\\nA means isolation >= 61.\\nD means emergency >= 10.\\nC means permit >= 61.\\nB means crew >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": 81, \\\"permit\\\": 67, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0096:noul", "group_id": "wf12:utility_maintenance:test:0096", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c4812edd45.\\nA means isolation >= 61.\\nD means emergency >= 10.\\nC means permit >= 61.\\nB means crew >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 83, \\\"crew\\\": 81, \\\"permit\\\": 67, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0096:uncertain", "group_id": "wf12:utility_maintenance:test:0096", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c4812edd45.\\nA means isolation >= 61.\\nD means emergency >= 10.\\nC means permit >= 61.\\nB means crew >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 81, \\\"permit\\\": 67, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.95, \"1\": 0.04500000000000004, \"2\": 0.0050000000000000044}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0097:choice", "group_id": "wf12:utility_maintenance:test:0097", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 44376bc1c0.\\nD means emergency >= 66.\\nA means isolation >= 36.\\nB means crew >= 19.\\nC means permit >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 77, \\\"crew\\\": 12, \\\"permit\\\": 62, \\\"emergency\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c7bd7f\": \"Outcome PROCEED.\", \"candidate_526a66\": \"Outcome REVIEW.\", \"candidate_83bdb8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c7bd7f\": 0.0, \"candidate_526a66\": 1.0, \"candidate_83bdb8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0097:score", "group_id": "wf12:utility_maintenance:test:0097", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 44376bc1c0.\\nD means emergency >= 66.\\nA means isolation >= 36.\\nB means crew >= 19.\\nC means permit >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 77, \\\"crew\\\": 12, \\\"permit\\\": 62, \\\"emergency\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0097:noul", "group_id": "wf12:utility_maintenance:test:0097", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 44376bc1c0.\\nD means emergency >= 66.\\nA means isolation >= 36.\\nB means crew >= 19.\\nC means permit >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 77, \\\"crew\\\": 12, \\\"permit\\\": 62, \\\"emergency\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0097:uncertain", "group_id": "wf12:utility_maintenance:test:0097", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 44376bc1c0.\\nD means emergency >= 66.\\nA means isolation >= 36.\\nB means crew >= 19.\\nC means permit >= 39.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 77, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 16}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.2, \"1\": 0.8, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0098:choice", "group_id": "wf12:utility_maintenance:test:0098", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2144e196df.\\nC means permit >= 20.\\nB means crew >= 74.\\nA means isolation >= 71.\\nD means emergency >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 14, \\\"crew\\\": 16, \\\"permit\\\": 12, \\\"emergency\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7d2760\": \"Outcome PROCEED.\", \"candidate_5e2b96\": \"Outcome REVIEW.\", \"candidate_30184b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7d2760\": 0.0, \"candidate_5e2b96\": 0.0, \"candidate_30184b\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0098:score", "group_id": "wf12:utility_maintenance:test:0098", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2144e196df.\\nC means permit >= 20.\\nB means crew >= 74.\\nA means isolation >= 71.\\nD means emergency >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 14, \\\"crew\\\": 16, \\\"permit\\\": 12, \\\"emergency\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0098:noul", "group_id": "wf12:utility_maintenance:test:0098", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2144e196df.\\nC means permit >= 20.\\nB means crew >= 74.\\nA means isolation >= 71.\\nD means emergency >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 14, \\\"crew\\\": 16, \\\"permit\\\": 12, \\\"emergency\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0098:uncertain", "group_id": "wf12:utility_maintenance:test:0098", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2144e196df.\\nC means permit >= 20.\\nB means crew >= 74.\\nA means isolation >= 71.\\nD means emergency >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 16, \\\"permit\\\": 12, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.8960000000000001, \"2\": 0.10399999999999987}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0099:choice", "group_id": "wf12:utility_maintenance:test:0099", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fbeecce3a7.\\nD means emergency >= 70.\\nB means crew >= 27.\\nC means permit >= 36.\\nA means isolation >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": 45, \\\"permit\\\": 25, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e25ce3\": \"Outcome PROCEED.\", \"candidate_ca4ed2\": \"Outcome REVIEW.\", \"candidate_404eb1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e25ce3\": 1.0, \"candidate_ca4ed2\": 0.0, \"candidate_404eb1\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0099:score", "group_id": "wf12:utility_maintenance:test:0099", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fbeecce3a7.\\nD means emergency >= 70.\\nB means crew >= 27.\\nC means permit >= 36.\\nA means isolation >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": 45, \\\"permit\\\": 25, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0099:noul", "group_id": "wf12:utility_maintenance:test:0099", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fbeecce3a7.\\nD means emergency >= 70.\\nB means crew >= 27.\\nC means permit >= 36.\\nA means isolation >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 79, \\\"crew\\\": 45, \\\"permit\\\": 25, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0099:uncertain", "group_id": "wf12:utility_maintenance:test:0099", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fbeecce3a7.\\nD means emergency >= 70.\\nB means crew >= 27.\\nC means permit >= 36.\\nA means isolation >= 26.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.95, \\\"C\\\": 0.95, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.09512500000000002, \"true\": 0.904875}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0100:choice", "group_id": "wf12:utility_maintenance:test:0100", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ac0c0fc26c.\\nD means emergency >= 44.\\nA means isolation >= 27.\\nC means permit >= 19.\\nB means crew >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 29, \\\"crew\\\": 6, \\\"permit\\\": 18, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ad7403\": \"Outcome PROCEED.\", \"candidate_9e1043\": \"Outcome REVIEW.\", \"candidate_500e89\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ad7403\": 0.0, \"candidate_9e1043\": 1.0, \"candidate_500e89\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0100:score", "group_id": "wf12:utility_maintenance:test:0100", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ac0c0fc26c.\\nD means emergency >= 44.\\nA means isolation >= 27.\\nC means permit >= 19.\\nB means crew >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 29, \\\"crew\\\": 6, \\\"permit\\\": 18, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0100:noul", "group_id": "wf12:utility_maintenance:test:0100", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ac0c0fc26c.\\nD means emergency >= 44.\\nA means isolation >= 27.\\nC means permit >= 19.\\nB means crew >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 29, \\\"crew\\\": 6, \\\"permit\\\": 18, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0100:uncertain", "group_id": "wf12:utility_maintenance:test:0100", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ac0c0fc26c.\\nD means emergency >= 44.\\nA means isolation >= 27.\\nC means permit >= 19.\\nB means crew >= 23.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.2, \\\"C\\\": 0.35, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.729625, \"true\": 0.270375}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0101:choice", "group_id": "wf12:utility_maintenance:test:0101", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 68b484264a.\\nB means crew >= 32.\\nA means isolation >= 63.\\nD means emergency >= 62.\\nC means permit >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 58, \\\"crew\\\": 43, \\\"permit\\\": 26, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f3f800\": \"Outcome PROCEED.\", \"candidate_448324\": \"Outcome REVIEW.\", \"candidate_55d0f0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f3f800\": 0.0, \"candidate_448324\": 0.0, \"candidate_55d0f0\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0101:score", "group_id": "wf12:utility_maintenance:test:0101", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 68b484264a.\\nB means crew >= 32.\\nA means isolation >= 63.\\nD means emergency >= 62.\\nC means permit >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 58, \\\"crew\\\": 43, \\\"permit\\\": 26, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0101:noul", "group_id": "wf12:utility_maintenance:test:0101", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 68b484264a.\\nB means crew >= 32.\\nA means isolation >= 63.\\nD means emergency >= 62.\\nC means permit >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 58, \\\"crew\\\": 43, \\\"permit\\\": 26, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0101:uncertain", "group_id": "wf12:utility_maintenance:test:0101", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 68b484264a.\\nB means crew >= 32.\\nA means isolation >= 63.\\nD means emergency >= 62.\\nC means permit >= 28.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 58, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 6}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.8, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0102:choice", "group_id": "wf12:utility_maintenance:test:0102", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b2ddddb5b.\\nA means isolation >= 35.\\nB means crew >= 39.\\nD means emergency >= 59.\\nC means permit >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 89, \\\"crew\\\": 62, \\\"permit\\\": 1, \\\"emergency\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_626bd4\": \"Outcome PROCEED.\", \"candidate_690375\": \"Outcome REVIEW.\", \"candidate_268907\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_626bd4\": 1.0, \"candidate_690375\": 0.0, \"candidate_268907\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0102:score", "group_id": "wf12:utility_maintenance:test:0102", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b2ddddb5b.\\nA means isolation >= 35.\\nB means crew >= 39.\\nD means emergency >= 59.\\nC means permit >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 89, \\\"crew\\\": 62, \\\"permit\\\": 1, \\\"emergency\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0102:noul", "group_id": "wf12:utility_maintenance:test:0102", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b2ddddb5b.\\nA means isolation >= 35.\\nB means crew >= 39.\\nD means emergency >= 59.\\nC means permit >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 89, \\\"crew\\\": 62, \\\"permit\\\": 1, \\\"emergency\\\": 8}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0102:uncertain", "group_id": "wf12:utility_maintenance:test:0102", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b2ddddb5b.\\nA means isolation >= 35.\\nB means crew >= 39.\\nD means emergency >= 59.\\nC means permit >= 17.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 1, \\\"emergency\\\": 8}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_626bd4\": \"Outcome PROCEED.\", \"candidate_690375\": \"Outcome REVIEW.\", \"candidate_268907\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_626bd4\": 0.0475, \"candidate_690375\": 0.0025000000000000022, \"candidate_268907\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0103:choice", "group_id": "wf12:utility_maintenance:test:0103", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference adf3e829c5.\\nA means isolation >= 59.\\nC means permit >= 35.\\nB means crew >= 87.\\nD means emergency >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 20, \\\"crew\\\": 93, \\\"permit\\\": 83, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e8304c\": \"Outcome PROCEED.\", \"candidate_e6e10c\": \"Outcome REVIEW.\", \"candidate_b703f7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e8304c\": 0.0, \"candidate_e6e10c\": 1.0, \"candidate_b703f7\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0103:score", "group_id": "wf12:utility_maintenance:test:0103", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference adf3e829c5.\\nA means isolation >= 59.\\nC means permit >= 35.\\nB means crew >= 87.\\nD means emergency >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 20, \\\"crew\\\": 93, \\\"permit\\\": 83, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0103:noul", "group_id": "wf12:utility_maintenance:test:0103", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference adf3e829c5.\\nA means isolation >= 59.\\nC means permit >= 35.\\nB means crew >= 87.\\nD means emergency >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 20, \\\"crew\\\": 93, \\\"permit\\\": 83, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0103:uncertain", "group_id": "wf12:utility_maintenance:test:0103", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference adf3e829c5.\\nA means isolation >= 59.\\nC means permit >= 35.\\nB means crew >= 87.\\nD means emergency >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 93, \\\"permit\\\": 83, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e8304c\": \"Outcome PROCEED.\", \"candidate_e6e10c\": \"Outcome REVIEW.\", \"candidate_b703f7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e8304c\": 0.2, \"candidate_e6e10c\": 0.52, \"candidate_b703f7\": 0.27999999999999997}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0104:choice", "group_id": "wf12:utility_maintenance:test:0104", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 61e77a0068.\\nA means isolation >= 85.\\nD means emergency >= 79.\\nB means crew >= 81.\\nC means permit >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": 79, \\\"permit\\\": 9, \\\"emergency\\\": 83}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a5078e\": \"Outcome PROCEED.\", \"candidate_f75587\": \"Outcome REVIEW.\", \"candidate_4769c1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a5078e\": 0.0, \"candidate_f75587\": 0.0, \"candidate_4769c1\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0104:score", "group_id": "wf12:utility_maintenance:test:0104", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 61e77a0068.\\nA means isolation >= 85.\\nD means emergency >= 79.\\nB means crew >= 81.\\nC means permit >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": 79, \\\"permit\\\": 9, \\\"emergency\\\": 83}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0104:noul", "group_id": "wf12:utility_maintenance:test:0104", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 61e77a0068.\\nA means isolation >= 85.\\nD means emergency >= 79.\\nB means crew >= 81.\\nC means permit >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": 79, \\\"permit\\\": 9, \\\"emergency\\\": 83}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0104:uncertain", "group_id": "wf12:utility_maintenance:test:0104", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 61e77a0068.\\nA means isolation >= 85.\\nD means emergency >= 79.\\nB means crew >= 81.\\nC means permit >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.8, \\\"C\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a5078e\": \"Outcome PROCEED.\", \"candidate_f75587\": \"Outcome REVIEW.\", \"candidate_4769c1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a5078e\": 0.7357280000000002, \"candidate_f75587\": 0.0887719999999999, \"candidate_4769c1\": 0.17549999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0105:choice", "group_id": "wf12:utility_maintenance:test:0105", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eae04a4220.\\nD means emergency >= 76.\\nB means crew >= 52.\\nC means permit >= 47.\\nA means isolation >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 93, \\\"crew\\\": 77, \\\"permit\\\": 67, \\\"emergency\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_04f120\": \"Outcome PROCEED.\", \"candidate_8e545c\": \"Outcome REVIEW.\", \"candidate_5ee847\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_04f120\": 1.0, \"candidate_8e545c\": 0.0, \"candidate_5ee847\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0105:score", "group_id": "wf12:utility_maintenance:test:0105", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eae04a4220.\\nD means emergency >= 76.\\nB means crew >= 52.\\nC means permit >= 47.\\nA means isolation >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 93, \\\"crew\\\": 77, \\\"permit\\\": 67, \\\"emergency\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0105:noul", "group_id": "wf12:utility_maintenance:test:0105", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eae04a4220.\\nD means emergency >= 76.\\nB means crew >= 52.\\nC means permit >= 47.\\nA means isolation >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 93, \\\"crew\\\": 77, \\\"permit\\\": 67, \\\"emergency\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0105:uncertain", "group_id": "wf12:utility_maintenance:test:0105", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference eae04a4220.\\nD means emergency >= 76.\\nB means crew >= 52.\\nC means permit >= 47.\\nA means isolation >= 62.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 77, \\\"permit\\\": 67, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.95, \"1\": 0.0025000000000000044, \"2\": 0.04750000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0106:choice", "group_id": "wf12:utility_maintenance:test:0106", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fa6f4e90a9.\\nC means permit >= 41.\\nB means crew >= 25.\\nD means emergency >= 12.\\nA means isolation >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 87, \\\"crew\\\": 13, \\\"permit\\\": 54, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9f4985\": \"Outcome PROCEED.\", \"candidate_d9ca5a\": \"Outcome REVIEW.\", \"candidate_909ee0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9f4985\": 0.0, \"candidate_d9ca5a\": 1.0, \"candidate_909ee0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0106:score", "group_id": "wf12:utility_maintenance:test:0106", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fa6f4e90a9.\\nC means permit >= 41.\\nB means crew >= 25.\\nD means emergency >= 12.\\nA means isolation >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 87, \\\"crew\\\": 13, \\\"permit\\\": 54, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0106:noul", "group_id": "wf12:utility_maintenance:test:0106", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fa6f4e90a9.\\nC means permit >= 41.\\nB means crew >= 25.\\nD means emergency >= 12.\\nA means isolation >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 87, \\\"crew\\\": 13, \\\"permit\\\": 54, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0106:uncertain", "group_id": "wf12:utility_maintenance:test:0106", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fa6f4e90a9.\\nC means permit >= 41.\\nB means crew >= 25.\\nD means emergency >= 12.\\nA means isolation >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 13, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 7}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"C\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.72, \"2\": 0.27999999999999997}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0107:choice", "group_id": "wf12:utility_maintenance:test:0107", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 36bde42fb9.\\nD means emergency >= 77.\\nC means permit >= 72.\\nA means isolation >= 46.\\nB means crew >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 34, \\\"crew\\\": 59, \\\"permit\\\": 97, \\\"emergency\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_18d9cc\": \"Outcome PROCEED.\", \"candidate_470b44\": \"Outcome REVIEW.\", \"candidate_175790\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_18d9cc\": 0.0, \"candidate_470b44\": 0.0, \"candidate_175790\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0107:score", "group_id": "wf12:utility_maintenance:test:0107", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 36bde42fb9.\\nD means emergency >= 77.\\nC means permit >= 72.\\nA means isolation >= 46.\\nB means crew >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 34, \\\"crew\\\": 59, \\\"permit\\\": 97, \\\"emergency\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0107:noul", "group_id": "wf12:utility_maintenance:test:0107", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 36bde42fb9.\\nD means emergency >= 77.\\nC means permit >= 72.\\nA means isolation >= 46.\\nB means crew >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 34, \\\"crew\\\": 59, \\\"permit\\\": 97, \\\"emergency\\\": 84}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0107:uncertain", "group_id": "wf12:utility_maintenance:test:0107", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 36bde42fb9.\\nD means emergency >= 77.\\nC means permit >= 72.\\nA means isolation >= 46.\\nB means crew >= 78.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 59, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 84}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"C\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.31360000000000005, \"1\": 0.5824000000000001, \"2\": 0.10399999999999987}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0108:choice", "group_id": "wf12:utility_maintenance:test:0108", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a2dd8c10df.\\nC means permit >= 31.\\nA means isolation >= 89.\\nB means crew >= 36.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 55, \\\"permit\\\": 22, \\\"emergency\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_099741\": \"Outcome PROCEED.\", \"candidate_f42de2\": \"Outcome REVIEW.\", \"candidate_25c830\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_099741\": 1.0, \"candidate_f42de2\": 0.0, \"candidate_25c830\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0108:score", "group_id": "wf12:utility_maintenance:test:0108", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a2dd8c10df.\\nC means permit >= 31.\\nA means isolation >= 89.\\nB means crew >= 36.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 55, \\\"permit\\\": 22, \\\"emergency\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0108:noul", "group_id": "wf12:utility_maintenance:test:0108", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a2dd8c10df.\\nC means permit >= 31.\\nA means isolation >= 89.\\nB means crew >= 36.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 55, \\\"permit\\\": 22, \\\"emergency\\\": 16}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0108:uncertain", "group_id": "wf12:utility_maintenance:test:0108", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a2dd8c10df.\\nC means permit >= 31.\\nA means isolation >= 89.\\nB means crew >= 36.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 55, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 16}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"C\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.995, \"true\": 0.005000000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0109:choice", "group_id": "wf12:utility_maintenance:test:0109", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6ddf61d737.\\nD means emergency >= 50.\\nB means crew >= 21.\\nA means isolation >= 78.\\nC means permit >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 32, \\\"crew\\\": 4, \\\"permit\\\": 79, \\\"emergency\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ea26a5\": \"Outcome PROCEED.\", \"candidate_836ce2\": \"Outcome REVIEW.\", \"candidate_44a8b7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ea26a5\": 0.0, \"candidate_836ce2\": 1.0, \"candidate_44a8b7\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0109:score", "group_id": "wf12:utility_maintenance:test:0109", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6ddf61d737.\\nD means emergency >= 50.\\nB means crew >= 21.\\nA means isolation >= 78.\\nC means permit >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 32, \\\"crew\\\": 4, \\\"permit\\\": 79, \\\"emergency\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0109:noul", "group_id": "wf12:utility_maintenance:test:0109", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6ddf61d737.\\nD means emergency >= 50.\\nB means crew >= 21.\\nA means isolation >= 78.\\nC means permit >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 32, \\\"crew\\\": 4, \\\"permit\\\": 79, \\\"emergency\\\": 42}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0109:uncertain", "group_id": "wf12:utility_maintenance:test:0109", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6ddf61d737.\\nD means emergency >= 50.\\nB means crew >= 21.\\nA means isolation >= 78.\\nC means permit >= 52.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 32, \\\"crew\\\": 4, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.15999999999999992, \"true\": 0.8400000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0110:choice", "group_id": "wf12:utility_maintenance:test:0110", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d9aded8df4.\\nD means emergency >= 75.\\nA means isolation >= 21.\\nC means permit >= 27.\\nB means crew >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 1, \\\"permit\\\": 22, \\\"emergency\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4a9012\": \"Outcome PROCEED.\", \"candidate_f312cc\": \"Outcome REVIEW.\", \"candidate_094573\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4a9012\": 0.0, \"candidate_f312cc\": 0.0, \"candidate_094573\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0110:score", "group_id": "wf12:utility_maintenance:test:0110", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d9aded8df4.\\nD means emergency >= 75.\\nA means isolation >= 21.\\nC means permit >= 27.\\nB means crew >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 1, \\\"permit\\\": 22, \\\"emergency\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0110:noul", "group_id": "wf12:utility_maintenance:test:0110", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d9aded8df4.\\nD means emergency >= 75.\\nA means isolation >= 21.\\nC means permit >= 27.\\nB means crew >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 1, \\\"permit\\\": 22, \\\"emergency\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0110:uncertain", "group_id": "wf12:utility_maintenance:test:0110", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d9aded8df4.\\nD means emergency >= 75.\\nA means isolation >= 21.\\nC means permit >= 27.\\nB means crew >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 1, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 50}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.96, \"true\": 0.03999999999999998}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0111:choice", "group_id": "wf12:utility_maintenance:test:0111", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 659491148f.\\nB means crew >= 89.\\nA means isolation >= 59.\\nD means emergency >= 32.\\nC means permit >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 90, \\\"crew\\\": 47, \\\"permit\\\": 74, \\\"emergency\\\": 70}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cb6c90\": \"Outcome PROCEED.\", \"candidate_cb55f9\": \"Outcome REVIEW.\", \"candidate_b1d42b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cb6c90\": 1.0, \"candidate_cb55f9\": 0.0, \"candidate_b1d42b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0111:score", "group_id": "wf12:utility_maintenance:test:0111", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 659491148f.\\nB means crew >= 89.\\nA means isolation >= 59.\\nD means emergency >= 32.\\nC means permit >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 90, \\\"crew\\\": 47, \\\"permit\\\": 74, \\\"emergency\\\": 70}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0111:noul", "group_id": "wf12:utility_maintenance:test:0111", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 659491148f.\\nB means crew >= 89.\\nA means isolation >= 59.\\nD means emergency >= 32.\\nC means permit >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 90, \\\"crew\\\": 47, \\\"permit\\\": 74, \\\"emergency\\\": 70}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0111:uncertain", "group_id": "wf12:utility_maintenance:test:0111", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 659491148f.\\nB means crew >= 89.\\nA means isolation >= 59.\\nD means emergency >= 32.\\nC means permit >= 63.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 90, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 70}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"C\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cb6c90\": \"Outcome PROCEED.\", \"candidate_cb55f9\": \"Outcome REVIEW.\", \"candidate_b1d42b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cb6c90\": 0.9525, \"candidate_cb55f9\": 0.04750000000000004, \"candidate_b1d42b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0112:choice", "group_id": "wf12:utility_maintenance:test:0112", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 57d374bddc.\\nC means permit >= 84.\\nB means crew >= 67.\\nD means emergency >= 63.\\nA means isolation >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 48, \\\"permit\\\": 89, \\\"emergency\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_51f760\": \"Outcome PROCEED.\", \"candidate_f7c5d4\": \"Outcome REVIEW.\", \"candidate_8897fc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_51f760\": 0.0, \"candidate_f7c5d4\": 1.0, \"candidate_8897fc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0112:score", "group_id": "wf12:utility_maintenance:test:0112", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 57d374bddc.\\nC means permit >= 84.\\nB means crew >= 67.\\nD means emergency >= 63.\\nA means isolation >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 48, \\\"permit\\\": 89, \\\"emergency\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0112:noul", "group_id": "wf12:utility_maintenance:test:0112", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 57d374bddc.\\nC means permit >= 84.\\nB means crew >= 67.\\nD means emergency >= 63.\\nA means isolation >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 48, \\\"permit\\\": 89, \\\"emergency\\\": 18}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0112:uncertain", "group_id": "wf12:utility_maintenance:test:0112", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 57d374bddc.\\nC means permit >= 84.\\nB means crew >= 67.\\nD means emergency >= 63.\\nA means isolation >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 18}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_51f760\": \"Outcome PROCEED.\", \"candidate_f7c5d4\": \"Outcome REVIEW.\", \"candidate_8897fc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_51f760\": 0.2, \"candidate_f7c5d4\": 0.8, \"candidate_8897fc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0113:choice", "group_id": "wf12:utility_maintenance:test:0113", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f4da769394.\\nA means isolation >= 89.\\nC means permit >= 65.\\nB means crew >= 38.\\nD means emergency >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 86, \\\"crew\\\": 1, \\\"permit\\\": 31, \\\"emergency\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c5ada2\": \"Outcome PROCEED.\", \"candidate_6b51d2\": \"Outcome REVIEW.\", \"candidate_8105b5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c5ada2\": 0.0, \"candidate_6b51d2\": 0.0, \"candidate_8105b5\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0113:score", "group_id": "wf12:utility_maintenance:test:0113", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f4da769394.\\nA means isolation >= 89.\\nC means permit >= 65.\\nB means crew >= 38.\\nD means emergency >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 86, \\\"crew\\\": 1, \\\"permit\\\": 31, \\\"emergency\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0113:noul", "group_id": "wf12:utility_maintenance:test:0113", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f4da769394.\\nA means isolation >= 89.\\nC means permit >= 65.\\nB means crew >= 38.\\nD means emergency >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 86, \\\"crew\\\": 1, \\\"permit\\\": 31, \\\"emergency\\\": 31}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0113:uncertain", "group_id": "wf12:utility_maintenance:test:0113", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f4da769394.\\nA means isolation >= 89.\\nC means permit >= 65.\\nB means crew >= 38.\\nD means emergency >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.2, \\\"C\\\": 0.8, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c5ada2\": \"Outcome PROCEED.\", \"candidate_6b51d2\": \"Outcome REVIEW.\", \"candidate_8105b5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c5ada2\": 0.4255999999999999, \"candidate_6b51d2\": 0.4784, \"candidate_8105b5\": 0.09599999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0114:choice", "group_id": "wf12:utility_maintenance:test:0114", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 367b66016e.\\nB means crew >= 19.\\nD means emergency >= 45.\\nC means permit >= 88.\\nA means isolation >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 39, \\\"crew\\\": 15, \\\"permit\\\": 93, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7a58d7\": \"Outcome PROCEED.\", \"candidate_864ff6\": \"Outcome REVIEW.\", \"candidate_750105\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7a58d7\": 1.0, \"candidate_864ff6\": 0.0, \"candidate_750105\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0114:score", "group_id": "wf12:utility_maintenance:test:0114", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 367b66016e.\\nB means crew >= 19.\\nD means emergency >= 45.\\nC means permit >= 88.\\nA means isolation >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 39, \\\"crew\\\": 15, \\\"permit\\\": 93, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0114:noul", "group_id": "wf12:utility_maintenance:test:0114", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 367b66016e.\\nB means crew >= 19.\\nD means emergency >= 45.\\nC means permit >= 88.\\nA means isolation >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 39, \\\"crew\\\": 15, \\\"permit\\\": 93, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0114:uncertain", "group_id": "wf12:utility_maintenance:test:0114", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 367b66016e.\\nB means crew >= 19.\\nD means emergency >= 45.\\nC means permit >= 88.\\nA means isolation >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 39, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 93, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.955, \"1\": 0.04500000000000004, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0115:choice", "group_id": "wf12:utility_maintenance:test:0115", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 42a060bb1a.\\nC means permit >= 44.\\nB means crew >= 32.\\nA means isolation >= 58.\\nD means emergency >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 42, \\\"crew\\\": 41, \\\"permit\\\": 83, \\\"emergency\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_73e02e\": \"Outcome PROCEED.\", \"candidate_da3378\": \"Outcome REVIEW.\", \"candidate_115854\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_73e02e\": 0.0, \"candidate_da3378\": 1.0, \"candidate_115854\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0115:score", "group_id": "wf12:utility_maintenance:test:0115", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 42a060bb1a.\\nC means permit >= 44.\\nB means crew >= 32.\\nA means isolation >= 58.\\nD means emergency >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 42, \\\"crew\\\": 41, \\\"permit\\\": 83, \\\"emergency\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0115:noul", "group_id": "wf12:utility_maintenance:test:0115", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 42a060bb1a.\\nC means permit >= 44.\\nB means crew >= 32.\\nA means isolation >= 58.\\nD means emergency >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 42, \\\"crew\\\": 41, \\\"permit\\\": 83, \\\"emergency\\\": 56}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0115:uncertain", "group_id": "wf12:utility_maintenance:test:0115", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 42a060bb1a.\\nC means permit >= 44.\\nB means crew >= 32.\\nA means isolation >= 58.\\nD means emergency >= 59.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 41, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 56}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"C\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.20000000000000004, \"1\": 0.16000000000000003, \"2\": 0.6400000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0116:choice", "group_id": "wf12:utility_maintenance:test:0116", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5eb6817494.\\nA means isolation >= 29.\\nB means crew >= 90.\\nD means emergency >= 28.\\nC means permit >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 15, \\\"crew\\\": 94, \\\"permit\\\": 92, \\\"emergency\\\": 54}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_decb62\": \"Outcome PROCEED.\", \"candidate_604b6d\": \"Outcome REVIEW.\", \"candidate_fecbe7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_decb62\": 0.0, \"candidate_604b6d\": 0.0, \"candidate_fecbe7\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0116:score", "group_id": "wf12:utility_maintenance:test:0116", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5eb6817494.\\nA means isolation >= 29.\\nB means crew >= 90.\\nD means emergency >= 28.\\nC means permit >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 15, \\\"crew\\\": 94, \\\"permit\\\": 92, \\\"emergency\\\": 54}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0116:noul", "group_id": "wf12:utility_maintenance:test:0116", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5eb6817494.\\nA means isolation >= 29.\\nB means crew >= 90.\\nD means emergency >= 28.\\nC means permit >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 15, \\\"crew\\\": 94, \\\"permit\\\": 92, \\\"emergency\\\": 54}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0116:uncertain", "group_id": "wf12:utility_maintenance:test:0116", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5eb6817494.\\nA means isolation >= 29.\\nB means crew >= 90.\\nD means emergency >= 28.\\nC means permit >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.2, \\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.2681, \"1\": 0.18589999999999995, \"2\": 0.546}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0117:choice", "group_id": "wf12:utility_maintenance:test:0117", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a3bdccefa7.\\nB means crew >= 83.\\nC means permit >= 56.\\nD means emergency >= 37.\\nA means isolation >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 76, \\\"crew\\\": 91, \\\"permit\\\": 22, \\\"emergency\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_03b358\": \"Outcome PROCEED.\", \"candidate_b12b50\": \"Outcome REVIEW.\", \"candidate_5723f7\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_03b358\": 1.0, \"candidate_b12b50\": 0.0, \"candidate_5723f7\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0117:score", "group_id": "wf12:utility_maintenance:test:0117", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a3bdccefa7.\\nB means crew >= 83.\\nC means permit >= 56.\\nD means emergency >= 37.\\nA means isolation >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 76, \\\"crew\\\": 91, \\\"permit\\\": 22, \\\"emergency\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0117:noul", "group_id": "wf12:utility_maintenance:test:0117", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a3bdccefa7.\\nB means crew >= 83.\\nC means permit >= 56.\\nD means emergency >= 37.\\nA means isolation >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 76, \\\"crew\\\": 91, \\\"permit\\\": 22, \\\"emergency\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0117:uncertain", "group_id": "wf12:utility_maintenance:test:0117", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference a3bdccefa7.\\nB means crew >= 83.\\nC means permit >= 56.\\nD means emergency >= 37.\\nA means isolation >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.1, \\\"C\\\": 0.9, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8694999999999999, \"true\": 0.1305}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0118:choice", "group_id": "wf12:utility_maintenance:test:0118", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c04a86f159.\\nC means permit >= 88.\\nD means emergency >= 37.\\nB means crew >= 14.\\nA means isolation >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 7, \\\"permit\\\": 93, \\\"emergency\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3bfd5f\": \"Outcome PROCEED.\", \"candidate_cabf47\": \"Outcome REVIEW.\", \"candidate_ccf265\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3bfd5f\": 0.0, \"candidate_cabf47\": 1.0, \"candidate_ccf265\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0118:score", "group_id": "wf12:utility_maintenance:test:0118", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c04a86f159.\\nC means permit >= 88.\\nD means emergency >= 37.\\nB means crew >= 14.\\nA means isolation >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 7, \\\"permit\\\": 93, \\\"emergency\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0118:noul", "group_id": "wf12:utility_maintenance:test:0118", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c04a86f159.\\nC means permit >= 88.\\nD means emergency >= 37.\\nB means crew >= 14.\\nA means isolation >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 7, \\\"permit\\\": 93, \\\"emergency\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0118:uncertain", "group_id": "wf12:utility_maintenance:test:0118", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c04a86f159.\\nC means permit >= 88.\\nD means emergency >= 37.\\nB means crew >= 14.\\nA means isolation >= 15.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.65, \\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8320000000000001, \"true\": 0.16799999999999995}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0119:choice", "group_id": "wf12:utility_maintenance:test:0119", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d58f6c0470.\\nD means emergency >= 34.\\nA means isolation >= 87.\\nB means crew >= 64.\\nC means permit >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 61, \\\"crew\\\": 54, \\\"permit\\\": 21, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d07a78\": \"Outcome PROCEED.\", \"candidate_d16d10\": \"Outcome REVIEW.\", \"candidate_c146ea\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d07a78\": 0.0, \"candidate_d16d10\": 0.0, \"candidate_c146ea\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0119:score", "group_id": "wf12:utility_maintenance:test:0119", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d58f6c0470.\\nD means emergency >= 34.\\nA means isolation >= 87.\\nB means crew >= 64.\\nC means permit >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 61, \\\"crew\\\": 54, \\\"permit\\\": 21, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0119:noul", "group_id": "wf12:utility_maintenance:test:0119", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d58f6c0470.\\nD means emergency >= 34.\\nA means isolation >= 87.\\nB means crew >= 64.\\nC means permit >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 61, \\\"crew\\\": 54, \\\"permit\\\": 21, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0119:uncertain", "group_id": "wf12:utility_maintenance:test:0119", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d58f6c0470.\\nD means emergency >= 34.\\nA means isolation >= 87.\\nB means crew >= 64.\\nC means permit >= 29.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 54, \\\"permit\\\": 21, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8, \"true\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0120:choice", "group_id": "wf12:utility_maintenance:test:0120", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ac92c2f47e.\\nB means crew >= 47.\\nA means isolation >= 85.\\nC means permit >= 31.\\nD means emergency >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": 54, \\\"permit\\\": 6, \\\"emergency\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4eb445\": \"Outcome PROCEED.\", \"candidate_d25f0b\": \"Outcome REVIEW.\", \"candidate_857f62\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4eb445\": 1.0, \"candidate_d25f0b\": 0.0, \"candidate_857f62\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0120:score", "group_id": "wf12:utility_maintenance:test:0120", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ac92c2f47e.\\nB means crew >= 47.\\nA means isolation >= 85.\\nC means permit >= 31.\\nD means emergency >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": 54, \\\"permit\\\": 6, \\\"emergency\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0120:noul", "group_id": "wf12:utility_maintenance:test:0120", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ac92c2f47e.\\nB means crew >= 47.\\nA means isolation >= 85.\\nC means permit >= 31.\\nD means emergency >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": 54, \\\"permit\\\": 6, \\\"emergency\\\": 81}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0120:uncertain", "group_id": "wf12:utility_maintenance:test:0120", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ac92c2f47e.\\nB means crew >= 47.\\nA means isolation >= 85.\\nC means permit >= 31.\\nD means emergency >= 66.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 81}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"C\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4eb445\": \"Outcome PROCEED.\", \"candidate_d25f0b\": \"Outcome REVIEW.\", \"candidate_857f62\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4eb445\": 0.955, \"candidate_d25f0b\": 0.04500000000000004, \"candidate_857f62\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0121:choice", "group_id": "wf12:utility_maintenance:test:0121", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 98b577b9b8.\\nB means crew >= 36.\\nC means permit >= 51.\\nA means isolation >= 56.\\nD means emergency >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": 14, \\\"permit\\\": 94, \\\"emergency\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8ce758\": \"Outcome PROCEED.\", \"candidate_0eee3a\": \"Outcome REVIEW.\", \"candidate_e0a3a0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8ce758\": 0.0, \"candidate_0eee3a\": 1.0, \"candidate_e0a3a0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0121:score", "group_id": "wf12:utility_maintenance:test:0121", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 98b577b9b8.\\nB means crew >= 36.\\nC means permit >= 51.\\nA means isolation >= 56.\\nD means emergency >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": 14, \\\"permit\\\": 94, \\\"emergency\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0121:noul", "group_id": "wf12:utility_maintenance:test:0121", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 98b577b9b8.\\nB means crew >= 36.\\nC means permit >= 51.\\nA means isolation >= 56.\\nD means emergency >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": 14, \\\"permit\\\": 94, \\\"emergency\\\": 29}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0121:uncertain", "group_id": "wf12:utility_maintenance:test:0121", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 98b577b9b8.\\nB means crew >= 36.\\nC means permit >= 51.\\nA means isolation >= 56.\\nD means emergency >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 85, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 29}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8ce758\": \"Outcome PROCEED.\", \"candidate_0eee3a\": \"Outcome REVIEW.\", \"candidate_e0a3a0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8ce758\": 0.35, \"candidate_0eee3a\": 0.65, \"candidate_e0a3a0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0122:choice", "group_id": "wf12:utility_maintenance:test:0122", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference dca312f853.\\nC means permit >= 76.\\nA means isolation >= 62.\\nB means crew >= 73.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 19, \\\"permit\\\": 5, \\\"emergency\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9d8678\": \"Outcome PROCEED.\", \"candidate_139acb\": \"Outcome REVIEW.\", \"candidate_58f6e6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9d8678\": 0.0, \"candidate_139acb\": 0.0, \"candidate_58f6e6\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0122:score", "group_id": "wf12:utility_maintenance:test:0122", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference dca312f853.\\nC means permit >= 76.\\nA means isolation >= 62.\\nB means crew >= 73.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 19, \\\"permit\\\": 5, \\\"emergency\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0122:noul", "group_id": "wf12:utility_maintenance:test:0122", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference dca312f853.\\nC means permit >= 76.\\nA means isolation >= 62.\\nB means crew >= 73.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 19, \\\"permit\\\": 5, \\\"emergency\\\": 67}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0122:uncertain", "group_id": "wf12:utility_maintenance:test:0122", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference dca312f853.\\nC means permit >= 76.\\nA means isolation >= 62.\\nB means crew >= 73.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 5, \\\"emergency\\\": 67}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9d8678\": \"Outcome PROCEED.\", \"candidate_139acb\": \"Outcome REVIEW.\", \"candidate_58f6e6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9d8678\": 0.42958904109589036, \"candidate_139acb\": 0.10739726027397256, \"candidate_58f6e6\": 0.4630136986301371}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0123:choice", "group_id": "wf12:utility_maintenance:test:0123", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 28d8c60119.\\nA means isolation >= 75.\\nD means emergency >= 35.\\nB means crew >= 28.\\nC means permit >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 87, \\\"permit\\\": 40, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5a8049\": \"Outcome PROCEED.\", \"candidate_ed04e8\": \"Outcome REVIEW.\", \"candidate_472b9f\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5a8049\": 1.0, \"candidate_ed04e8\": 0.0, \"candidate_472b9f\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0123:score", "group_id": "wf12:utility_maintenance:test:0123", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 28d8c60119.\\nA means isolation >= 75.\\nD means emergency >= 35.\\nB means crew >= 28.\\nC means permit >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 87, \\\"permit\\\": 40, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0123:noul", "group_id": "wf12:utility_maintenance:test:0123", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 28d8c60119.\\nA means isolation >= 75.\\nD means emergency >= 35.\\nB means crew >= 28.\\nC means permit >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 88, \\\"crew\\\": 87, \\\"permit\\\": 40, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0123:uncertain", "group_id": "wf12:utility_maintenance:test:0123", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 28d8c60119.\\nA means isolation >= 75.\\nD means emergency >= 35.\\nB means crew >= 28.\\nC means permit >= 53.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.95, \\\"C\\\": 0.1, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.09502500000000001, \"1\": 0.09047500000000001, \"2\": 0.8145}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0124:choice", "group_id": "wf12:utility_maintenance:test:0124", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference bc755aba8d.\\nD means emergency >= 34.\\nC means permit >= 79.\\nA means isolation >= 60.\\nB means crew >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 71, \\\"crew\\\": 51, \\\"permit\\\": 88, \\\"emergency\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c9a680\": \"Outcome PROCEED.\", \"candidate_735c6b\": \"Outcome REVIEW.\", \"candidate_2def34\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c9a680\": 0.0, \"candidate_735c6b\": 1.0, \"candidate_2def34\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0124:score", "group_id": "wf12:utility_maintenance:test:0124", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference bc755aba8d.\\nD means emergency >= 34.\\nC means permit >= 79.\\nA means isolation >= 60.\\nB means crew >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 71, \\\"crew\\\": 51, \\\"permit\\\": 88, \\\"emergency\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0124:noul", "group_id": "wf12:utility_maintenance:test:0124", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference bc755aba8d.\\nD means emergency >= 34.\\nC means permit >= 79.\\nA means isolation >= 60.\\nB means crew >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 71, \\\"crew\\\": 51, \\\"permit\\\": 88, \\\"emergency\\\": 21}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0124:uncertain", "group_id": "wf12:utility_maintenance:test:0124", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference bc755aba8d.\\nD means emergency >= 34.\\nC means permit >= 79.\\nA means isolation >= 60.\\nB means crew >= 70.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 71, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 21}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"C\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.65, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0125:choice", "group_id": "wf12:utility_maintenance:test:0125", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 17386e725f.\\nB means crew >= 48.\\nA means isolation >= 85.\\nC means permit >= 14.\\nD means emergency >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 78, \\\"permit\\\": 44, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f0cf21\": \"Outcome PROCEED.\", \"candidate_31cc90\": \"Outcome REVIEW.\", \"candidate_5ce539\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f0cf21\": 0.0, \"candidate_31cc90\": 0.0, \"candidate_5ce539\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0125:score", "group_id": "wf12:utility_maintenance:test:0125", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 17386e725f.\\nB means crew >= 48.\\nA means isolation >= 85.\\nC means permit >= 14.\\nD means emergency >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 78, \\\"permit\\\": 44, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0125:noul", "group_id": "wf12:utility_maintenance:test:0125", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 17386e725f.\\nB means crew >= 48.\\nA means isolation >= 85.\\nC means permit >= 14.\\nD means emergency >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 11, \\\"crew\\\": 78, \\\"permit\\\": 44, \\\"emergency\\\": 69}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0125:uncertain", "group_id": "wf12:utility_maintenance:test:0125", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 17386e725f.\\nB means crew >= 48.\\nA means isolation >= 85.\\nC means permit >= 14.\\nD means emergency >= 58.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 44, \\\"emergency\\\": 69}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.5369863013698629, \"1\": 0.0, \"2\": 0.4630136986301371}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0126:choice", "group_id": "wf12:utility_maintenance:test:0126", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d4983f366d.\\nB means crew >= 80.\\nC means permit >= 14.\\nA means isolation >= 14.\\nD means emergency >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 40, \\\"crew\\\": 96, \\\"permit\\\": 4, \\\"emergency\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_29c5cd\": \"Outcome PROCEED.\", \"candidate_83a6a5\": \"Outcome REVIEW.\", \"candidate_fb2f97\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_29c5cd\": 1.0, \"candidate_83a6a5\": 0.0, \"candidate_fb2f97\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0126:score", "group_id": "wf12:utility_maintenance:test:0126", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d4983f366d.\\nB means crew >= 80.\\nC means permit >= 14.\\nA means isolation >= 14.\\nD means emergency >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 40, \\\"crew\\\": 96, \\\"permit\\\": 4, \\\"emergency\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0126:noul", "group_id": "wf12:utility_maintenance:test:0126", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d4983f366d.\\nB means crew >= 80.\\nC means permit >= 14.\\nA means isolation >= 14.\\nD means emergency >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 40, \\\"crew\\\": 96, \\\"permit\\\": 4, \\\"emergency\\\": 0}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0126:uncertain", "group_id": "wf12:utility_maintenance:test:0126", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference d4983f366d.\\nB means crew >= 80.\\nC means permit >= 14.\\nA means isolation >= 14.\\nD means emergency >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.1, \\\"B\\\": 0.9, \\\"C\\\": 0.05, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.10225000000000006, \"true\": 0.8977499999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0127:choice", "group_id": "wf12:utility_maintenance:test:0127", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd42bdf941.\\nD means emergency >= 69.\\nB means crew >= 18.\\nC means permit >= 88.\\nA means isolation >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 8, \\\"permit\\\": 61, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d46304\": \"Outcome PROCEED.\", \"candidate_1c20a5\": \"Outcome REVIEW.\", \"candidate_f6c4c9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d46304\": 0.0, \"candidate_1c20a5\": 1.0, \"candidate_f6c4c9\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0127:score", "group_id": "wf12:utility_maintenance:test:0127", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd42bdf941.\\nD means emergency >= 69.\\nB means crew >= 18.\\nC means permit >= 88.\\nA means isolation >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 8, \\\"permit\\\": 61, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0127:noul", "group_id": "wf12:utility_maintenance:test:0127", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd42bdf941.\\nD means emergency >= 69.\\nB means crew >= 18.\\nC means permit >= 88.\\nA means isolation >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 97, \\\"crew\\\": 8, \\\"permit\\\": 61, \\\"emergency\\\": 88}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0127:uncertain", "group_id": "wf12:utility_maintenance:test:0127", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd42bdf941.\\nD means emergency >= 69.\\nB means crew >= 18.\\nC means permit >= 88.\\nA means isolation >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 8, \\\"permit\\\": 61, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8, \"true\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0128:choice", "group_id": "wf12:utility_maintenance:test:0128", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3f27da7e8c.\\nB means crew >= 20.\\nD means emergency >= 20.\\nC means permit >= 16.\\nA means isolation >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": 100, \\\"permit\\\": 70, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_26e31d\": \"Outcome PROCEED.\", \"candidate_18b49e\": \"Outcome REVIEW.\", \"candidate_140b1c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_26e31d\": 0.0, \"candidate_18b49e\": 0.0, \"candidate_140b1c\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0128:score", "group_id": "wf12:utility_maintenance:test:0128", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3f27da7e8c.\\nB means crew >= 20.\\nD means emergency >= 20.\\nC means permit >= 16.\\nA means isolation >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": 100, \\\"permit\\\": 70, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0128:noul", "group_id": "wf12:utility_maintenance:test:0128", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3f27da7e8c.\\nB means crew >= 20.\\nD means emergency >= 20.\\nC means permit >= 16.\\nA means isolation >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": 100, \\\"permit\\\": 70, \\\"emergency\\\": 94}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0128:uncertain", "group_id": "wf12:utility_maintenance:test:0128", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 3f27da7e8c.\\nB means crew >= 20.\\nD means emergency >= 20.\\nC means permit >= 16.\\nA means isolation >= 18.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 100, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 94}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"C\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8, \"true\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0129:choice", "group_id": "wf12:utility_maintenance:test:0129", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference aa74801ed8.\\nD means emergency >= 28.\\nC means permit >= 44.\\nA means isolation >= 47.\\nB means crew >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 76, \\\"permit\\\": 21, \\\"emergency\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f35907\": \"Outcome PROCEED.\", \"candidate_7ebad1\": \"Outcome REVIEW.\", \"candidate_bc194c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f35907\": 1.0, \"candidate_7ebad1\": 0.0, \"candidate_bc194c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0129:score", "group_id": "wf12:utility_maintenance:test:0129", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference aa74801ed8.\\nD means emergency >= 28.\\nC means permit >= 44.\\nA means isolation >= 47.\\nB means crew >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 76, \\\"permit\\\": 21, \\\"emergency\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0129:noul", "group_id": "wf12:utility_maintenance:test:0129", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference aa74801ed8.\\nD means emergency >= 28.\\nC means permit >= 44.\\nA means isolation >= 47.\\nB means crew >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 82, \\\"crew\\\": 76, \\\"permit\\\": 21, \\\"emergency\\\": 41}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0129:uncertain", "group_id": "wf12:utility_maintenance:test:0129", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference aa74801ed8.\\nD means emergency >= 28.\\nC means permit >= 44.\\nA means isolation >= 47.\\nB means crew >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 21, \\\"emergency\\\": 41}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.05, \\\"B\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f35907\": \"Outcome PROCEED.\", \"candidate_7ebad1\": \"Outcome REVIEW.\", \"candidate_bc194c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f35907\": 0.005000000000000001, \"candidate_7ebad1\": 0.045000000000000005, \"candidate_bc194c\": 0.95}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0130:choice", "group_id": "wf12:utility_maintenance:test:0130", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2524115172.\\nC means permit >= 46.\\nA means isolation >= 52.\\nB means crew >= 27.\\nD means emergency >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 5, \\\"permit\\\": 5, \\\"emergency\\\": 26}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_440f4e\": \"Outcome PROCEED.\", \"candidate_d06069\": \"Outcome REVIEW.\", \"candidate_507d50\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_440f4e\": 0.0, \"candidate_d06069\": 1.0, \"candidate_507d50\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0130:score", "group_id": "wf12:utility_maintenance:test:0130", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2524115172.\\nC means permit >= 46.\\nA means isolation >= 52.\\nB means crew >= 27.\\nD means emergency >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 5, \\\"permit\\\": 5, \\\"emergency\\\": 26}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0130:noul", "group_id": "wf12:utility_maintenance:test:0130", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2524115172.\\nC means permit >= 46.\\nA means isolation >= 52.\\nB means crew >= 27.\\nD means emergency >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 5, \\\"permit\\\": 5, \\\"emergency\\\": 26}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0130:uncertain", "group_id": "wf12:utility_maintenance:test:0130", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2524115172.\\nC means permit >= 46.\\nA means isolation >= 52.\\nB means crew >= 27.\\nD means emergency >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 55, \\\"crew\\\": 5, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_440f4e\": \"Outcome PROCEED.\", \"candidate_d06069\": \"Outcome REVIEW.\", \"candidate_507d50\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_440f4e\": 0.42250000000000004, \"candidate_d06069\": 0.5774999999999999, \"candidate_507d50\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0131:choice", "group_id": "wf12:utility_maintenance:test:0131", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cacb7f57d6.\\nB means crew >= 10.\\nC means permit >= 23.\\nA means isolation >= 48.\\nD means emergency >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 36, \\\"crew\\\": 31, \\\"permit\\\": 10, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f9863f\": \"Outcome PROCEED.\", \"candidate_f414ae\": \"Outcome REVIEW.\", \"candidate_b531bc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f9863f\": 0.0, \"candidate_f414ae\": 0.0, \"candidate_b531bc\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0131:score", "group_id": "wf12:utility_maintenance:test:0131", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cacb7f57d6.\\nB means crew >= 10.\\nC means permit >= 23.\\nA means isolation >= 48.\\nD means emergency >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 36, \\\"crew\\\": 31, \\\"permit\\\": 10, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0131:noul", "group_id": "wf12:utility_maintenance:test:0131", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cacb7f57d6.\\nB means crew >= 10.\\nC means permit >= 23.\\nA means isolation >= 48.\\nD means emergency >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 36, \\\"crew\\\": 31, \\\"permit\\\": 10, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0131:uncertain", "group_id": "wf12:utility_maintenance:test:0131", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference cacb7f57d6.\\nB means crew >= 10.\\nC means permit >= 23.\\nA means isolation >= 48.\\nD means emergency >= 51.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 36, \\\"crew\\\": 31, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.8, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f9863f\": \"Outcome PROCEED.\", \"candidate_f414ae\": \"Outcome REVIEW.\", \"candidate_b531bc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f9863f\": 0.0, \"candidate_f414ae\": 0.1792, \"candidate_b531bc\": 0.8208}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0132:choice", "group_id": "wf12:utility_maintenance:test:0132", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f4fb9fb75b.\\nA means isolation >= 77.\\nD means emergency >= 83.\\nC means permit >= 40.\\nB means crew >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 81, \\\"crew\\\": 79, \\\"permit\\\": 40, \\\"emergency\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_26dbba\": \"Outcome PROCEED.\", \"candidate_97368e\": \"Outcome REVIEW.\", \"candidate_3a7683\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_26dbba\": 1.0, \"candidate_97368e\": 0.0, \"candidate_3a7683\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0132:score", "group_id": "wf12:utility_maintenance:test:0132", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f4fb9fb75b.\\nA means isolation >= 77.\\nD means emergency >= 83.\\nC means permit >= 40.\\nB means crew >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 81, \\\"crew\\\": 79, \\\"permit\\\": 40, \\\"emergency\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0132:noul", "group_id": "wf12:utility_maintenance:test:0132", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f4fb9fb75b.\\nA means isolation >= 77.\\nD means emergency >= 83.\\nC means permit >= 40.\\nB means crew >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 81, \\\"crew\\\": 79, \\\"permit\\\": 40, \\\"emergency\\\": 95}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0132:uncertain", "group_id": "wf12:utility_maintenance:test:0132", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f4fb9fb75b.\\nA means isolation >= 77.\\nD means emergency >= 83.\\nC means permit >= 40.\\nB means crew >= 69.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 81, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 40, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.05, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.9525, \"1\": 0.04750000000000004, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0133:choice", "group_id": "wf12:utility_maintenance:test:0133", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 417e944ca4.\\nB means crew >= 71.\\nC means permit >= 26.\\nA means isolation >= 41.\\nD means emergency >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 14, \\\"crew\\\": 99, \\\"permit\\\": 40, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_26b565\": \"Outcome PROCEED.\", \"candidate_d4646e\": \"Outcome REVIEW.\", \"candidate_8e8969\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_26b565\": 0.0, \"candidate_d4646e\": 1.0, \"candidate_8e8969\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0133:score", "group_id": "wf12:utility_maintenance:test:0133", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 417e944ca4.\\nB means crew >= 71.\\nC means permit >= 26.\\nA means isolation >= 41.\\nD means emergency >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 14, \\\"crew\\\": 99, \\\"permit\\\": 40, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0133:noul", "group_id": "wf12:utility_maintenance:test:0133", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 417e944ca4.\\nB means crew >= 71.\\nC means permit >= 26.\\nA means isolation >= 41.\\nD means emergency >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 14, \\\"crew\\\": 99, \\\"permit\\\": 40, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0133:uncertain", "group_id": "wf12:utility_maintenance:test:0133", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 417e944ca4.\\nB means crew >= 71.\\nC means permit >= 26.\\nA means isolation >= 41.\\nD means emergency >= 22.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 99, \\\"permit\\\": 40, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.22749999999999998, \"2\": 0.42250000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0134:choice", "group_id": "wf12:utility_maintenance:test:0134", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 04ca49f30c.\\nD means emergency >= 77.\\nC means permit >= 23.\\nB means crew >= 66.\\nA means isolation >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 30, \\\"crew\\\": 90, \\\"permit\\\": 48, \\\"emergency\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_064f01\": \"Outcome PROCEED.\", \"candidate_9e0c03\": \"Outcome REVIEW.\", \"candidate_9e9dcc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_064f01\": 0.0, \"candidate_9e0c03\": 0.0, \"candidate_9e9dcc\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0134:score", "group_id": "wf12:utility_maintenance:test:0134", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 04ca49f30c.\\nD means emergency >= 77.\\nC means permit >= 23.\\nB means crew >= 66.\\nA means isolation >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 30, \\\"crew\\\": 90, \\\"permit\\\": 48, \\\"emergency\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0134:noul", "group_id": "wf12:utility_maintenance:test:0134", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 04ca49f30c.\\nD means emergency >= 77.\\nC means permit >= 23.\\nB means crew >= 66.\\nA means isolation >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 30, \\\"crew\\\": 90, \\\"permit\\\": 48, \\\"emergency\\\": 82}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0134:uncertain", "group_id": "wf12:utility_maintenance:test:0134", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 04ca49f30c.\\nD means emergency >= 77.\\nC means permit >= 23.\\nB means crew >= 66.\\nA means isolation >= 74.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 90, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 82}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.5369863013698629, \"1\": 0.0, \"2\": 0.4630136986301371}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0135:choice", "group_id": "wf12:utility_maintenance:test:0135", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b11c28df7.\\nD means emergency >= 59.\\nA means isolation >= 77.\\nC means permit >= 86.\\nB means crew >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 87, \\\"crew\\\": 64, \\\"permit\\\": 98, \\\"emergency\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_931a11\": \"Outcome PROCEED.\", \"candidate_47ad54\": \"Outcome REVIEW.\", \"candidate_08c84d\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_931a11\": 1.0, \"candidate_47ad54\": 0.0, \"candidate_08c84d\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0135:score", "group_id": "wf12:utility_maintenance:test:0135", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b11c28df7.\\nD means emergency >= 59.\\nA means isolation >= 77.\\nC means permit >= 86.\\nB means crew >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 87, \\\"crew\\\": 64, \\\"permit\\\": 98, \\\"emergency\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0135:noul", "group_id": "wf12:utility_maintenance:test:0135", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b11c28df7.\\nD means emergency >= 59.\\nA means isolation >= 77.\\nC means permit >= 86.\\nB means crew >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 87, \\\"crew\\\": 64, \\\"permit\\\": 98, \\\"emergency\\\": 60}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0135:uncertain", "group_id": "wf12:utility_maintenance:test:0135", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2b11c28df7.\\nD means emergency >= 59.\\nA means isolation >= 77.\\nC means permit >= 86.\\nB means crew >= 57.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.9, \\\"C\\\": 0.1, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9544999999999999, \"true\": 0.04550000000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0136:choice", "group_id": "wf12:utility_maintenance:test:0136", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 187ef42850.\\nD means emergency >= 50.\\nA means isolation >= 28.\\nC means permit >= 48.\\nB means crew >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 4, \\\"permit\\\": 35, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_548cad\": \"Outcome PROCEED.\", \"candidate_529f61\": \"Outcome REVIEW.\", \"candidate_112bdc\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_548cad\": 0.0, \"candidate_529f61\": 1.0, \"candidate_112bdc\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0136:score", "group_id": "wf12:utility_maintenance:test:0136", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 187ef42850.\\nD means emergency >= 50.\\nA means isolation >= 28.\\nC means permit >= 48.\\nB means crew >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 4, \\\"permit\\\": 35, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0136:noul", "group_id": "wf12:utility_maintenance:test:0136", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 187ef42850.\\nD means emergency >= 50.\\nA means isolation >= 28.\\nC means permit >= 48.\\nB means crew >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 4, \\\"permit\\\": 35, \\\"emergency\\\": 57}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0136:uncertain", "group_id": "wf12:utility_maintenance:test:0136", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 187ef42850.\\nD means emergency >= 50.\\nA means isolation >= 28.\\nC means permit >= 48.\\nB means crew >= 13.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 4, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 57}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"C\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.35, \"true\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0137:choice", "group_id": "wf12:utility_maintenance:test:0137", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 96c8d0240a.\\nB means crew >= 44.\\nA means isolation >= 16.\\nC means permit >= 25.\\nD means emergency >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 72, \\\"permit\\\": 17, \\\"emergency\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4dac73\": \"Outcome PROCEED.\", \"candidate_f8f41a\": \"Outcome REVIEW.\", \"candidate_b13898\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4dac73\": 0.0, \"candidate_f8f41a\": 0.0, \"candidate_b13898\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0137:score", "group_id": "wf12:utility_maintenance:test:0137", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 96c8d0240a.\\nB means crew >= 44.\\nA means isolation >= 16.\\nC means permit >= 25.\\nD means emergency >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 72, \\\"permit\\\": 17, \\\"emergency\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0137:noul", "group_id": "wf12:utility_maintenance:test:0137", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 96c8d0240a.\\nB means crew >= 44.\\nA means isolation >= 16.\\nC means permit >= 25.\\nD means emergency >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 13, \\\"crew\\\": 72, \\\"permit\\\": 17, \\\"emergency\\\": 50}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0137:uncertain", "group_id": "wf12:utility_maintenance:test:0137", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 96c8d0240a.\\nB means crew >= 44.\\nA means isolation >= 16.\\nC means permit >= 25.\\nD means emergency >= 71.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.8, \\\"B\\\": 0.2, \\\"C\\\": 0.35, \\\"D\\\": 0.8}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.8140000000000001, \"true\": 0.18599999999999994}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0138:choice", "group_id": "wf12:utility_maintenance:test:0138", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e84ac9369b.\\nB means crew >= 14.\\nA means isolation >= 43.\\nC means permit >= 61.\\nD means emergency >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 54, \\\"crew\\\": 95, \\\"permit\\\": 70, \\\"emergency\\\": 34}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_08c327\": \"Outcome PROCEED.\", \"candidate_9e8300\": \"Outcome REVIEW.\", \"candidate_6c38e9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_08c327\": 1.0, \"candidate_9e8300\": 0.0, \"candidate_6c38e9\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0138:score", "group_id": "wf12:utility_maintenance:test:0138", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e84ac9369b.\\nB means crew >= 14.\\nA means isolation >= 43.\\nC means permit >= 61.\\nD means emergency >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 54, \\\"crew\\\": 95, \\\"permit\\\": 70, \\\"emergency\\\": 34}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0138:noul", "group_id": "wf12:utility_maintenance:test:0138", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e84ac9369b.\\nB means crew >= 14.\\nA means isolation >= 43.\\nC means permit >= 61.\\nD means emergency >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 54, \\\"crew\\\": 95, \\\"permit\\\": 70, \\\"emergency\\\": 34}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0138:uncertain", "group_id": "wf12:utility_maintenance:test:0138", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e84ac9369b.\\nB means crew >= 14.\\nA means isolation >= 43.\\nC means permit >= 61.\\nD means emergency >= 14.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 54, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 70, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.9, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_08c327\": \"Outcome PROCEED.\", \"candidate_9e8300\": \"Outcome REVIEW.\", \"candidate_6c38e9\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_08c327\": 0.99, \"candidate_9e8300\": 0.009999999999999995, \"candidate_6c38e9\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0139:choice", "group_id": "wf12:utility_maintenance:test:0139", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 540878fd06.\\nB means crew >= 28.\\nC means permit >= 82.\\nA means isolation >= 48.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 15, \\\"permit\\\": 86, \\\"emergency\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_462c27\": \"Outcome PROCEED.\", \"candidate_a89632\": \"Outcome REVIEW.\", \"candidate_611389\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_462c27\": 0.0, \"candidate_a89632\": 1.0, \"candidate_611389\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0139:score", "group_id": "wf12:utility_maintenance:test:0139", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 540878fd06.\\nB means crew >= 28.\\nC means permit >= 82.\\nA means isolation >= 48.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 15, \\\"permit\\\": 86, \\\"emergency\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0139:noul", "group_id": "wf12:utility_maintenance:test:0139", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 540878fd06.\\nB means crew >= 28.\\nC means permit >= 82.\\nA means isolation >= 48.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": 15, \\\"permit\\\": 86, \\\"emergency\\\": 17}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0139:uncertain", "group_id": "wf12:utility_maintenance:test:0139", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 540878fd06.\\nB means crew >= 28.\\nC means permit >= 82.\\nA means isolation >= 48.\\nD means emergency >= 42.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 45, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 17}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"C\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_462c27\": \"Outcome PROCEED.\", \"candidate_a89632\": \"Outcome REVIEW.\", \"candidate_611389\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_462c27\": 0.0, \"candidate_a89632\": 0.8, \"candidate_611389\": 0.19999999999999996}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0140:choice", "group_id": "wf12:utility_maintenance:test:0140", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 986574ae60.\\nD means emergency >= 36.\\nA means isolation >= 20.\\nC means permit >= 41.\\nB means crew >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": 56, \\\"permit\\\": 85, \\\"emergency\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ac0e84\": \"Outcome PROCEED.\", \"candidate_b5b91b\": \"Outcome REVIEW.\", \"candidate_6d04f1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ac0e84\": 0.0, \"candidate_b5b91b\": 0.0, \"candidate_6d04f1\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0140:score", "group_id": "wf12:utility_maintenance:test:0140", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 986574ae60.\\nD means emergency >= 36.\\nA means isolation >= 20.\\nC means permit >= 41.\\nB means crew >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": 56, \\\"permit\\\": 85, \\\"emergency\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0140:noul", "group_id": "wf12:utility_maintenance:test:0140", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 986574ae60.\\nD means emergency >= 36.\\nA means isolation >= 20.\\nC means permit >= 41.\\nB means crew >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": 56, \\\"permit\\\": 85, \\\"emergency\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0140:uncertain", "group_id": "wf12:utility_maintenance:test:0140", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 986574ae60.\\nD means emergency >= 36.\\nA means isolation >= 20.\\nC means permit >= 41.\\nB means crew >= 48.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 1, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 85, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ac0e84\": \"Outcome PROCEED.\", \"candidate_b5b91b\": \"Outcome REVIEW.\", \"candidate_6d04f1\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ac0e84\": 0.0, \"candidate_b5b91b\": 0.8, \"candidate_6d04f1\": 0.2}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0141:choice", "group_id": "wf12:utility_maintenance:test:0141", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f473b18551.\\nC means permit >= 10.\\nA means isolation >= 25.\\nD means emergency >= 49.\\nB means crew >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 66, \\\"crew\\\": 88, \\\"permit\\\": 7, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2c328f\": \"Outcome PROCEED.\", \"candidate_43c630\": \"Outcome REVIEW.\", \"candidate_524da5\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2c328f\": 1.0, \"candidate_43c630\": 0.0, \"candidate_524da5\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0141:score", "group_id": "wf12:utility_maintenance:test:0141", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f473b18551.\\nC means permit >= 10.\\nA means isolation >= 25.\\nD means emergency >= 49.\\nB means crew >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 66, \\\"crew\\\": 88, \\\"permit\\\": 7, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0141:noul", "group_id": "wf12:utility_maintenance:test:0141", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f473b18551.\\nC means permit >= 10.\\nA means isolation >= 25.\\nD means emergency >= 49.\\nB means crew >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 66, \\\"crew\\\": 88, \\\"permit\\\": 7, \\\"emergency\\\": 99}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0141:uncertain", "group_id": "wf12:utility_maintenance:test:0141", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f473b18551.\\nC means permit >= 10.\\nA means isolation >= 25.\\nD means emergency >= 49.\\nB means crew >= 86.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 66, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 99}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"C\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.19000000000000003, \"1\": 0.81, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0142:choice", "group_id": "wf12:utility_maintenance:test:0142", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6fc93dce5e.\\nD means emergency >= 10.\\nA means isolation >= 55.\\nB means crew >= 43.\\nC means permit >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 63, \\\"crew\\\": 34, \\\"permit\\\": 96, \\\"emergency\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ba6edb\": \"Outcome PROCEED.\", \"candidate_3118bc\": \"Outcome REVIEW.\", \"candidate_a6aa20\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ba6edb\": 0.0, \"candidate_3118bc\": 1.0, \"candidate_a6aa20\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0142:score", "group_id": "wf12:utility_maintenance:test:0142", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6fc93dce5e.\\nD means emergency >= 10.\\nA means isolation >= 55.\\nB means crew >= 43.\\nC means permit >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 63, \\\"crew\\\": 34, \\\"permit\\\": 96, \\\"emergency\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0142:noul", "group_id": "wf12:utility_maintenance:test:0142", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6fc93dce5e.\\nD means emergency >= 10.\\nA means isolation >= 55.\\nB means crew >= 43.\\nC means permit >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 63, \\\"crew\\\": 34, \\\"permit\\\": 96, \\\"emergency\\\": 4}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0142:uncertain", "group_id": "wf12:utility_maintenance:test:0142", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 6fc93dce5e.\\nD means emergency >= 10.\\nA means isolation >= 55.\\nB means crew >= 43.\\nC means permit >= 87.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 63, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 96, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.48, \"1\": 0.52, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0143:choice", "group_id": "wf12:utility_maintenance:test:0143", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f034aaa0b9.\\nA means isolation >= 79.\\nC means permit >= 13.\\nB means crew >= 72.\\nD means emergency >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 16, \\\"permit\\\": 0, \\\"emergency\\\": 34}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c65244\": \"Outcome PROCEED.\", \"candidate_230de4\": \"Outcome REVIEW.\", \"candidate_5bfffb\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c65244\": 0.0, \"candidate_230de4\": 0.0, \"candidate_5bfffb\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0143:score", "group_id": "wf12:utility_maintenance:test:0143", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f034aaa0b9.\\nA means isolation >= 79.\\nC means permit >= 13.\\nB means crew >= 72.\\nD means emergency >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 16, \\\"permit\\\": 0, \\\"emergency\\\": 34}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0143:noul", "group_id": "wf12:utility_maintenance:test:0143", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f034aaa0b9.\\nA means isolation >= 79.\\nC means permit >= 13.\\nB means crew >= 72.\\nD means emergency >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 16, \\\"permit\\\": 0, \\\"emergency\\\": 34}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0143:uncertain", "group_id": "wf12:utility_maintenance:test:0143", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference f034aaa0b9.\\nA means isolation >= 79.\\nC means permit >= 13.\\nB means crew >= 72.\\nD means emergency >= 31.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 16, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.52, \"2\": 0.4799999999999999}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0144:choice", "group_id": "wf12:utility_maintenance:test:0144", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 168864d086.\\nB means crew >= 63.\\nC means permit >= 56.\\nA means isolation >= 74.\\nD means emergency >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": 74, \\\"permit\\\": 45, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_814d6f\": \"Outcome PROCEED.\", \"candidate_387a9f\": \"Outcome REVIEW.\", \"candidate_c072da\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_814d6f\": 1.0, \"candidate_387a9f\": 0.0, \"candidate_c072da\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0144:score", "group_id": "wf12:utility_maintenance:test:0144", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 168864d086.\\nB means crew >= 63.\\nC means permit >= 56.\\nA means isolation >= 74.\\nD means emergency >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": 74, \\\"permit\\\": 45, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0144:noul", "group_id": "wf12:utility_maintenance:test:0144", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 168864d086.\\nB means crew >= 63.\\nC means permit >= 56.\\nA means isolation >= 74.\\nD means emergency >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 95, \\\"crew\\\": 74, \\\"permit\\\": 45, \\\"emergency\\\": 2}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0144:uncertain", "group_id": "wf12:utility_maintenance:test:0144", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 168864d086.\\nB means crew >= 63.\\nC means permit >= 56.\\nA means isolation >= 74.\\nD means emergency >= 16.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.9, \\\"C\\\": 0.05, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.952375, \"true\": 0.04762500000000004}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0145:choice", "group_id": "wf12:utility_maintenance:test:0145", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 06f895e161.\\nD means emergency >= 42.\\nB means crew >= 33.\\nA means isolation >= 59.\\nC means permit >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 7, \\\"permit\\\": 18, \\\"emergency\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0c6671\": \"Outcome PROCEED.\", \"candidate_b5254a\": \"Outcome REVIEW.\", \"candidate_495d39\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0c6671\": 0.0, \"candidate_b5254a\": 1.0, \"candidate_495d39\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0145:score", "group_id": "wf12:utility_maintenance:test:0145", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 06f895e161.\\nD means emergency >= 42.\\nB means crew >= 33.\\nA means isolation >= 59.\\nC means permit >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 7, \\\"permit\\\": 18, \\\"emergency\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0145:noul", "group_id": "wf12:utility_maintenance:test:0145", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 06f895e161.\\nD means emergency >= 42.\\nB means crew >= 33.\\nA means isolation >= 59.\\nC means permit >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 98, \\\"crew\\\": 7, \\\"permit\\\": 18, \\\"emergency\\\": 3}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0145:uncertain", "group_id": "wf12:utility_maintenance:test:0145", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 06f895e161.\\nD means emergency >= 42.\\nB means crew >= 33.\\nA means isolation >= 59.\\nC means permit >= 43.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.2, \\\"C\\\": 0.2, \\\"D\\\": 0.65}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.2559999999999998, \"true\": 0.7440000000000002}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0146:choice", "group_id": "wf12:utility_maintenance:test:0146", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fbb3d0e25e.\\nA means isolation >= 51.\\nB means crew >= 47.\\nC means permit >= 66.\\nD means emergency >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 34, \\\"crew\\\": 13, \\\"permit\\\": 50, \\\"emergency\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1457a8\": \"Outcome PROCEED.\", \"candidate_2c2e19\": \"Outcome REVIEW.\", \"candidate_3237b6\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1457a8\": 0.0, \"candidate_2c2e19\": 0.0, \"candidate_3237b6\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0146:score", "group_id": "wf12:utility_maintenance:test:0146", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fbb3d0e25e.\\nA means isolation >= 51.\\nB means crew >= 47.\\nC means permit >= 66.\\nD means emergency >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 34, \\\"crew\\\": 13, \\\"permit\\\": 50, \\\"emergency\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0146:noul", "group_id": "wf12:utility_maintenance:test:0146", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fbb3d0e25e.\\nA means isolation >= 51.\\nB means crew >= 47.\\nC means permit >= 66.\\nD means emergency >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 34, \\\"crew\\\": 13, \\\"permit\\\": 50, \\\"emergency\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0146:uncertain", "group_id": "wf12:utility_maintenance:test:0146", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fbb3d0e25e.\\nA means isolation >= 51.\\nB means crew >= 47.\\nC means permit >= 66.\\nD means emergency >= 37.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.35, \\\"C\\\": 0.2, \\\"D\\\": 0.35}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.6955, \"true\": 0.30450000000000005}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0147:choice", "group_id": "wf12:utility_maintenance:test:0147", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 88ba926bc7.\\nA means isolation >= 20.\\nB means crew >= 62.\\nD means emergency >= 26.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 46, \\\"crew\\\": 81, \\\"permit\\\": 70, \\\"emergency\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4241fb\": \"Outcome PROCEED.\", \"candidate_83c462\": \"Outcome REVIEW.\", \"candidate_3462b4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4241fb\": 1.0, \"candidate_83c462\": 0.0, \"candidate_3462b4\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0147:score", "group_id": "wf12:utility_maintenance:test:0147", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 88ba926bc7.\\nA means isolation >= 20.\\nB means crew >= 62.\\nD means emergency >= 26.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 46, \\\"crew\\\": 81, \\\"permit\\\": 70, \\\"emergency\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0147:noul", "group_id": "wf12:utility_maintenance:test:0147", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 88ba926bc7.\\nA means isolation >= 20.\\nB means crew >= 62.\\nD means emergency >= 26.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 46, \\\"crew\\\": 81, \\\"permit\\\": 70, \\\"emergency\\\": 91}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0147:uncertain", "group_id": "wf12:utility_maintenance:test:0147", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 88ba926bc7.\\nA means isolation >= 20.\\nB means crew >= 62.\\nD means emergency >= 26.\\nC means permit >= 61.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": 81, \\\"permit\\\": 70, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"D\\\": 0.05}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4241fb\": \"Outcome PROCEED.\", \"candidate_83c462\": \"Outcome REVIEW.\", \"candidate_3462b4\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4241fb\": 0.95, \"candidate_83c462\": 0.04750000000000004, \"candidate_3462b4\": 0.0025000000000000022}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0148:choice", "group_id": "wf12:utility_maintenance:test:0148", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c771e2d5f9.\\nA means isolation >= 72.\\nB means crew >= 21.\\nC means permit >= 61.\\nD means emergency >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 20, \\\"permit\\\": 0, \\\"emergency\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c9f171\": \"Outcome PROCEED.\", \"candidate_c1d133\": \"Outcome REVIEW.\", \"candidate_e424a8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c9f171\": 0.0, \"candidate_c1d133\": 1.0, \"candidate_e424a8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0148:score", "group_id": "wf12:utility_maintenance:test:0148", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c771e2d5f9.\\nA means isolation >= 72.\\nB means crew >= 21.\\nC means permit >= 61.\\nD means emergency >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 20, \\\"permit\\\": 0, \\\"emergency\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0148:noul", "group_id": "wf12:utility_maintenance:test:0148", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c771e2d5f9.\\nA means isolation >= 72.\\nB means crew >= 21.\\nC means permit >= 61.\\nD means emergency >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 20, \\\"permit\\\": 0, \\\"emergency\\\": 71}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0148:uncertain", "group_id": "wf12:utility_maintenance:test:0148", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c771e2d5f9.\\nA means isolation >= 72.\\nB means crew >= 21.\\nC means permit >= 61.\\nD means emergency >= 10.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 80, \\\"crew\\\": 20, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c9f171\": \"Outcome PROCEED.\", \"candidate_c1d133\": \"Outcome REVIEW.\", \"candidate_e424a8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c9f171\": 0.06999999999999999, \"candidate_c1d133\": 0.9299999999999999, \"candidate_e424a8\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0149:choice", "group_id": "wf12:utility_maintenance:test:0149", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference acf5380189.\\nA means isolation >= 20.\\nD means emergency >= 43.\\nC means permit >= 54.\\nB means crew >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 9, \\\"crew\\\": 24, \\\"permit\\\": 34, \\\"emergency\\\": 54}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b7f69b\": \"Outcome PROCEED.\", \"candidate_b54b82\": \"Outcome REVIEW.\", \"candidate_35561c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b7f69b\": 0.0, \"candidate_b54b82\": 0.0, \"candidate_35561c\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0149:score", "group_id": "wf12:utility_maintenance:test:0149", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference acf5380189.\\nA means isolation >= 20.\\nD means emergency >= 43.\\nC means permit >= 54.\\nB means crew >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 9, \\\"crew\\\": 24, \\\"permit\\\": 34, \\\"emergency\\\": 54}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0149:noul", "group_id": "wf12:utility_maintenance:test:0149", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference acf5380189.\\nA means isolation >= 20.\\nD means emergency >= 43.\\nC means permit >= 54.\\nB means crew >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 9, \\\"crew\\\": 24, \\\"permit\\\": 34, \\\"emergency\\\": 54}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0149:uncertain", "group_id": "wf12:utility_maintenance:test:0149", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference acf5380189.\\nA means isolation >= 20.\\nD means emergency >= 43.\\nC means permit >= 54.\\nB means crew >= 79.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.35, \\\"C\\\": 0.2, \\\"D\\\": 0.65}. All unreported information remains unknown.\\nTwo independent sensors report on condition B. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b7f69b\": \"Outcome PROCEED.\", \"candidate_b54b82\": \"Outcome REVIEW.\", \"candidate_35561c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b7f69b\": 0.11943561643835615, \"candidate_b54b82\": 0.13656438356164385, \"candidate_35561c\": 0.7440000000000001}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0150:choice", "group_id": "wf12:utility_maintenance:test:0150", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 658a382839.\\nD means emergency >= 12.\\nA means isolation >= 62.\\nB means crew >= 78.\\nC means permit >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 87, \\\"crew\\\": 97, \\\"permit\\\": 77, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_183704\": \"Outcome PROCEED.\", \"candidate_a4de65\": \"Outcome REVIEW.\", \"candidate_5823e0\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_183704\": 1.0, \"candidate_a4de65\": 0.0, \"candidate_5823e0\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0150:score", "group_id": "wf12:utility_maintenance:test:0150", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 658a382839.\\nD means emergency >= 12.\\nA means isolation >= 62.\\nB means crew >= 78.\\nC means permit >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 87, \\\"crew\\\": 97, \\\"permit\\\": 77, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0150:noul", "group_id": "wf12:utility_maintenance:test:0150", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 658a382839.\\nD means emergency >= 12.\\nA means isolation >= 62.\\nB means crew >= 78.\\nC means permit >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 87, \\\"crew\\\": 97, \\\"permit\\\": 77, \\\"emergency\\\": 6}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0150:uncertain", "group_id": "wf12:utility_maintenance:test:0150", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 658a382839.\\nD means emergency >= 12.\\nA means isolation >= 62.\\nB means crew >= 78.\\nC means permit >= 76.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 77, \\\"emergency\\\": 6}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.095, \"1\": 0.905, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0151:choice", "group_id": "wf12:utility_maintenance:test:0151", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2ebab8f1e9.\\nC means permit >= 58.\\nD means emergency >= 27.\\nB means crew >= 24.\\nA means isolation >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": 17, \\\"permit\\\": 32, \\\"emergency\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e727d0\": \"Outcome PROCEED.\", \"candidate_8c432f\": \"Outcome REVIEW.\", \"candidate_a6baeb\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e727d0\": 0.0, \"candidate_8c432f\": 1.0, \"candidate_a6baeb\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0151:score", "group_id": "wf12:utility_maintenance:test:0151", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2ebab8f1e9.\\nC means permit >= 58.\\nD means emergency >= 27.\\nB means crew >= 24.\\nA means isolation >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": 17, \\\"permit\\\": 32, \\\"emergency\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0151:noul", "group_id": "wf12:utility_maintenance:test:0151", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2ebab8f1e9.\\nC means permit >= 58.\\nD means emergency >= 27.\\nB means crew >= 24.\\nA means isolation >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": 17, \\\"permit\\\": 32, \\\"emergency\\\": 14}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0151:uncertain", "group_id": "wf12:utility_maintenance:test:0151", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 2ebab8f1e9.\\nC means permit >= 58.\\nD means emergency >= 27.\\nB means crew >= 24.\\nA means isolation >= 85.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 32, \\\"emergency\\\": 14}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.35, \\\"B\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.27999999999999997, \"1\": 0.06999999999999998, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0152:choice", "group_id": "wf12:utility_maintenance:test:0152", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e9c6513620.\\nB means crew >= 62.\\nA means isolation >= 88.\\nD means emergency >= 13.\\nC means permit >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 26, \\\"crew\\\": 69, \\\"permit\\\": 44, \\\"emergency\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8234bb\": \"Outcome PROCEED.\", \"candidate_7850d7\": \"Outcome REVIEW.\", \"candidate_f2e7d8\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8234bb\": 0.0, \"candidate_7850d7\": 0.0, \"candidate_f2e7d8\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0152:score", "group_id": "wf12:utility_maintenance:test:0152", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e9c6513620.\\nB means crew >= 62.\\nA means isolation >= 88.\\nD means emergency >= 13.\\nC means permit >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 26, \\\"crew\\\": 69, \\\"permit\\\": 44, \\\"emergency\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0152:noul", "group_id": "wf12:utility_maintenance:test:0152", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e9c6513620.\\nB means crew >= 62.\\nA means isolation >= 88.\\nD means emergency >= 13.\\nC means permit >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 26, \\\"crew\\\": 69, \\\"permit\\\": 44, \\\"emergency\\\": 100}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0152:uncertain", "group_id": "wf12:utility_maintenance:test:0152", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference e9c6513620.\\nB means crew >= 62.\\nA means isolation >= 88.\\nD means emergency >= 13.\\nC means permit >= 32.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 44, \\\"emergency\\\": 100}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.2, \\\"B\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition A. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.35, \"1\": 0.0, \"2\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0153:choice", "group_id": "wf12:utility_maintenance:test:0153", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 990b2a6256.\\nC means permit >= 33.\\nB means crew >= 71.\\nA means isolation >= 24.\\nD means emergency >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": 77, \\\"permit\\\": 31, \\\"emergency\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ee5ba3\": \"Outcome PROCEED.\", \"candidate_28b6de\": \"Outcome REVIEW.\", \"candidate_26a168\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ee5ba3\": 1.0, \"candidate_28b6de\": 0.0, \"candidate_26a168\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0153:score", "group_id": "wf12:utility_maintenance:test:0153", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 990b2a6256.\\nC means permit >= 33.\\nB means crew >= 71.\\nA means isolation >= 24.\\nD means emergency >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": 77, \\\"permit\\\": 31, \\\"emergency\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0153:noul", "group_id": "wf12:utility_maintenance:test:0153", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 990b2a6256.\\nC means permit >= 33.\\nB means crew >= 71.\\nA means isolation >= 24.\\nD means emergency >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 60, \\\"crew\\\": 77, \\\"permit\\\": 31, \\\"emergency\\\": 61}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0153:uncertain", "group_id": "wf12:utility_maintenance:test:0153", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 990b2a6256.\\nC means permit >= 33.\\nB means crew >= 71.\\nA means isolation >= 24.\\nD means emergency >= 19.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.95, \\\"B\\\": 0.95, \\\"C\\\": 0.1, \\\"D\\\": 0.9}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.9504999999999999, \"true\": 0.049500000000000044}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0154:choice", "group_id": "wf12:utility_maintenance:test:0154", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 293bff89d2.\\nB means crew >= 10.\\nD means emergency >= 42.\\nA means isolation >= 59.\\nC means permit >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 73, \\\"crew\\\": 8, \\\"permit\\\": 79, \\\"emergency\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5eca57\": \"Outcome PROCEED.\", \"candidate_ce23d8\": \"Outcome REVIEW.\", \"candidate_7c290c\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5eca57\": 0.0, \"candidate_ce23d8\": 1.0, \"candidate_7c290c\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0154:score", "group_id": "wf12:utility_maintenance:test:0154", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 293bff89d2.\\nB means crew >= 10.\\nD means emergency >= 42.\\nA means isolation >= 59.\\nC means permit >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 73, \\\"crew\\\": 8, \\\"permit\\\": 79, \\\"emergency\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0154:noul", "group_id": "wf12:utility_maintenance:test:0154", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 293bff89d2.\\nB means crew >= 10.\\nD means emergency >= 42.\\nA means isolation >= 59.\\nC means permit >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 73, \\\"crew\\\": 8, \\\"permit\\\": 79, \\\"emergency\\\": 37}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0154:uncertain", "group_id": "wf12:utility_maintenance:test:0154", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 293bff89d2.\\nB means crew >= 10.\\nD means emergency >= 42.\\nA means isolation >= 59.\\nC means permit >= 21.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": \\\"unobserved\\\", \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"A\\\": 0.65, \\\"B\\\": 0.35, \\\"C\\\": 0.2, \\\"D\\\": 0.8}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.6639999999999999, \"true\": 0.336}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0155:choice", "group_id": "wf12:utility_maintenance:test:0155", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5c756d5327.\\nB means crew >= 54.\\nA means isolation >= 74.\\nD means emergency >= 81.\\nC means permit >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 67, \\\"permit\\\": 18, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_78dd40\": \"Outcome PROCEED.\", \"candidate_af8154\": \"Outcome REVIEW.\", \"candidate_d4f199\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_78dd40\": 0.0, \"candidate_af8154\": 0.0, \"candidate_d4f199\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0155:score", "group_id": "wf12:utility_maintenance:test:0155", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5c756d5327.\\nB means crew >= 54.\\nA means isolation >= 74.\\nD means emergency >= 81.\\nC means permit >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 67, \\\"permit\\\": 18, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0155:noul", "group_id": "wf12:utility_maintenance:test:0155", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5c756d5327.\\nB means crew >= 54.\\nA means isolation >= 74.\\nD means emergency >= 81.\\nC means permit >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 67, \\\"permit\\\": 18, \\\"emergency\\\": 98}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0155:uncertain", "group_id": "wf12:utility_maintenance:test:0155", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference 5c756d5327.\\nB means crew >= 54.\\nA means isolation >= 74.\\nD means emergency >= 81.\\nC means permit >= 60.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 6, \\\"crew\\\": 67, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.65, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules and stated uncertainty, what is the probability that HOLD is required?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.64, \"true\": 0.36}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0156:choice", "group_id": "wf12:utility_maintenance:test:0156", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c31524373b.\\nD means emergency >= 60.\\nC means permit >= 33.\\nB means crew >= 55.\\nA means isolation >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": 96, \\\"permit\\\": 4, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9fdeb3\": \"Outcome PROCEED.\", \"candidate_337c4f\": \"Outcome REVIEW.\", \"candidate_f30155\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9fdeb3\": 1.0, \"candidate_337c4f\": 0.0, \"candidate_f30155\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0156:score", "group_id": "wf12:utility_maintenance:test:0156", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c31524373b.\\nD means emergency >= 60.\\nC means permit >= 33.\\nB means crew >= 55.\\nA means isolation >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": 96, \\\"permit\\\": 4, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0156:noul", "group_id": "wf12:utility_maintenance:test:0156", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c31524373b.\\nD means emergency >= 60.\\nC means permit >= 33.\\nB means crew >= 55.\\nA means isolation >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": 96, \\\"permit\\\": 4, \\\"emergency\\\": 7}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0156:uncertain", "group_id": "wf12:utility_maintenance:test:0156", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference c31524373b.\\nD means emergency >= 60.\\nC means permit >= 33.\\nB means crew >= 55.\\nA means isolation >= 24.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 100, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 4, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.95, \\\"D\\\": 0.95}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9fdeb3\": \"Outcome PROCEED.\", \"candidate_337c4f\": \"Outcome REVIEW.\", \"candidate_f30155\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9fdeb3\": 0.95, \"candidate_337c4f\": 0.050000000000000044, \"candidate_f30155\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0157:choice", "group_id": "wf12:utility_maintenance:test:0157", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference b881694a75.\\nA means isolation >= 76.\\nD means emergency >= 40.\\nC means permit >= 10.\\nB means crew >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 12, \\\"crew\\\": 82, \\\"permit\\\": 88, \\\"emergency\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9ab0d6\": \"Outcome PROCEED.\", \"candidate_82d372\": \"Outcome REVIEW.\", \"candidate_67349b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9ab0d6\": 0.0, \"candidate_82d372\": 1.0, \"candidate_67349b\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0157:score", "group_id": "wf12:utility_maintenance:test:0157", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference b881694a75.\\nA means isolation >= 76.\\nD means emergency >= 40.\\nC means permit >= 10.\\nB means crew >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 12, \\\"crew\\\": 82, \\\"permit\\\": 88, \\\"emergency\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 1.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0157:noul", "group_id": "wf12:utility_maintenance:test:0157", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference b881694a75.\\nA means isolation >= 76.\\nD means emergency >= 40.\\nC means permit >= 10.\\nB means crew >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 12, \\\"crew\\\": 82, \\\"permit\\\": 88, \\\"emergency\\\": 22}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0157:uncertain", "group_id": "wf12:utility_maintenance:test:0157", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference b881694a75.\\nA means isolation >= 76.\\nD means emergency >= 40.\\nC means permit >= 10.\\nB means crew >= 54.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 12, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": 22}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.65, \\\"C\\\": 0.35}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9ab0d6\": \"Outcome PROCEED.\", \"candidate_82d372\": \"Outcome REVIEW.\", \"candidate_67349b\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9ab0d6\": 0.0, \"candidate_82d372\": 0.35, \"candidate_67349b\": 0.65}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0158:choice", "group_id": "wf12:utility_maintenance:test:0158", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ffec9bbe57.\\nC means permit >= 51.\\nD means emergency >= 75.\\nB means crew >= 55.\\nA means isolation >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 27, \\\"crew\\\": 42, \\\"permit\\\": 86, \\\"emergency\\\": 83}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_28eec0\": \"Outcome PROCEED.\", \"candidate_a695fa\": \"Outcome REVIEW.\", \"candidate_b513eb\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_28eec0\": 0.0, \"candidate_a695fa\": 0.0, \"candidate_b513eb\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0158:score", "group_id": "wf12:utility_maintenance:test:0158", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ffec9bbe57.\\nC means permit >= 51.\\nD means emergency >= 75.\\nB means crew >= 55.\\nA means isolation >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 27, \\\"crew\\\": 42, \\\"permit\\\": 86, \\\"emergency\\\": 83}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.0, \"1\": 0.0, \"2\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0158:noul", "group_id": "wf12:utility_maintenance:test:0158", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ffec9bbe57.\\nC means permit >= 51.\\nD means emergency >= 75.\\nB means crew >= 55.\\nA means isolation >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 27, \\\"crew\\\": 42, \\\"permit\\\": 86, \\\"emergency\\\": 83}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 0.0, \"true\": 1.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0158:uncertain", "group_id": "wf12:utility_maintenance:test:0158", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference ffec9bbe57.\\nC means permit >= 51.\\nD means emergency >= 75.\\nB means crew >= 55.\\nA means isolation >= 89.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 27, \\\"crew\\\": 42, \\\"permit\\\": \\\"unobserved\\\", \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"C\\\": 0.35, \\\"D\\\": 0.2}. All unreported information remains unknown.\\nTwo independent sensors report on condition C. Sensor X says TRUE and is correct with probability 0.80. Sensor Y says FALSE and is correct with probability 0.65. Each sensor has equal sensitivity and specificity. Their errors are independent conditional on the true condition. Use the reports and prior together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_28eec0\": \"Outcome PROCEED.\", \"candidate_a695fa\": \"Outcome REVIEW.\", \"candidate_b513eb\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_28eec0\": 0.0, \"candidate_a695fa\": 0.42958904109589036, \"candidate_b513eb\": 0.5704109589041096}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0159:choice", "group_id": "wf12:utility_maintenance:test:0159", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd41a728d0.\\nC means permit >= 90.\\nA means isolation >= 17.\\nD means emergency >= 77.\\nB means crew >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 49, \\\"crew\\\": 73, \\\"permit\\\": 94, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_69c7ef\": \"Outcome PROCEED.\", \"candidate_8ba975\": \"Outcome REVIEW.\", \"candidate_1ceb70\": \"Outcome HOLD.\"}}}}", "targets_json": "{\"decision\": {\"candidate_69c7ef\": 1.0, \"candidate_8ba975\": 0.0, \"candidate_1ceb70\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0159:score", "group_id": "wf12:utility_maintenance:test:0159", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd41a728d0.\\nC means permit >= 90.\\nA means isolation >= 17.\\nD means emergency >= 77.\\nB means crew >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 49, \\\"crew\\\": 73, \\\"permit\\\": 94, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 1.0, \"1\": 0.0, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0159:noul", "group_id": "wf12:utility_maintenance:test:0159", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd41a728d0.\\nC means permit >= 90.\\nA means isolation >= 17.\\nD means emergency >= 77.\\nB means crew >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 49, \\\"crew\\\": 73, \\\"permit\\\": 94, \\\"emergency\\\": 85}\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"noul\", \"instructions\": \"Under the supplied rules, is HOLD the required outcome?\", \"criteria\": {\"true\": \"The proposition is true.\", \"false\": \"The proposition is false.\"}}}}", "targets_json": "{\"decision\": {\"false\": 1.0, \"true\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "wf12:utility_maintenance:test:0159:uncertain", "group_id": "wf12:utility_maintenance:test:0159", "split": "test", "family": "utility_maintenance", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Workflow: utility maintenance. Case reference fd41a728d0.\\nC means permit >= 90.\\nA means isolation >= 17.\\nD means emergency >= 77.\\nB means crew >= 47.\\nApply the first matching rule in the listed priority order.\\n1. IF (NOT A AND (D OR NOT C)) THEN HOLD.\\n2. IF (A AND (B OR (C AND D))) THEN PROCEED.\\nOtherwise REVIEW.\\nObserved measurements: {\\\"isolation\\\": 49, \\\"crew\\\": \\\"unobserved\\\", \\\"permit\\\": 94, \\\"emergency\\\": \\\"unobserved\\\"}\\nFor this question, unknown conditions are independent Bernoulli variables with these specified prior probabilities: {\\\"B\\\": 0.1, \\\"D\\\": 0.1}. All unreported information remains unknown.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"score\", \"instructions\": \"Apply the supplied rules and return the distribution over the ordered outcomes.\", \"criteria\": [\"Outcome PROCEED.\", \"Outcome REVIEW.\", \"Outcome HOLD.\"]}}}", "targets_json": "{\"decision\": {\"0\": 0.19000000000000003, \"1\": 0.81, \"2\": 0.0}}", "source_id": "veyra-workflow-v12", "source_revision": "sha256:5433c424ad7d2bea710c398899840d4a43a3608a9b7529a6cd8cbbe7a32bc4ca", "source_license": "Apache-2.0", "source_url": "https://github.com/ken-jo/veyra", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2a280d72a06f84f5107fe612:choice", "group_id": "fresh12:snli:2a280d72a06f84f5107fe612", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Children playing a video game together.\\nHypothesis: The kids are indoors playing a video game.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_86320c\": \"The hypothesis follows from the premise.\", \"candidate_dc6edf\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_69fa04\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_86320c\": 1.0, \"candidate_dc6edf\": 0.0, \"candidate_69fa04\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d32511b5d41bac8dd9ec1f81:choice", "group_id": "fresh12:snli:d32511b5d41bac8dd9ec1f81", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a pink shirt and white vest plays a guitar.\\nHypothesis: There is a man wearing a pink shirt.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_45fef0\": \"The hypothesis follows from the premise.\", \"candidate_7fc2f6\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a6b4a7\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_45fef0\": 1.0, \"candidate_7fc2f6\": 0.0, \"candidate_a6b4a7\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d01900ce8ce554c0c348df5d:choice", "group_id": "fresh12:snli:d01900ce8ce554c0c348df5d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Little kids play a game of running around a pole.\\nHypothesis: There are kids outside playing\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8941aa\": \"The hypothesis follows from the premise.\", \"candidate_735e74\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f56850\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8941aa\": 1.0, \"candidate_735e74\": 0.0, \"candidate_f56850\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:21932fcfbbdeca629a1d49e6:choice", "group_id": "fresh12:snli:21932fcfbbdeca629a1d49e6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Man with tattoo is lying down in the sand next to a cow.\\nHypothesis: The man is outdoors and with animals.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_74e5b4\": \"The hypothesis follows from the premise.\", \"candidate_8bd7eb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_6a9c99\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_74e5b4\": 1.0, \"candidate_8bd7eb\": 0.0, \"candidate_6a9c99\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1f3c903f2c8ffa823540b37c:choice", "group_id": "fresh12:snli:1f3c903f2c8ffa823540b37c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Some kids are wrestling on an inflatable raft\\nHypothesis: Some kids playing with a raft.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dc10f9\": \"The hypothesis follows from the premise.\", \"candidate_00757d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1b3ea1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dc10f9\": 1.0, \"candidate_00757d\": 0.0, \"candidate_1b3ea1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4e833f81203607e4fbfc9a01:choice", "group_id": "fresh12:snli:4e833f81203607e4fbfc9a01", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young man blew up balloons to craft into animals for the seven excited children that looked on.\\nHypothesis: The children watch the man make dogs and giraffes out of balloons\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d52b2a\": \"The hypothesis follows from the premise.\", \"candidate_440c8b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_627f8e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d52b2a\": 1.0, \"candidate_440c8b\": 0.0, \"candidate_627f8e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d8f9f40184147eab89bdbb18:choice", "group_id": "fresh12:snli:d8f9f40184147eab89bdbb18", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A dog hides in the tall grasses along a rocky shore\\nHypothesis: The dog is outdoors.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_851f64\": \"The hypothesis follows from the premise.\", \"candidate_96fdc7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f3e85b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_851f64\": 1.0, \"candidate_96fdc7\": 0.0, \"candidate_f3e85b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a487271e73d1a5fa0c7315c3:choice", "group_id": "fresh12:snli:a487271e73d1a5fa0c7315c3", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of footballs players is standing behind a coaching official.\\nHypothesis: There are multiple people present.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a4917f\": \"The hypothesis follows from the premise.\", \"candidate_d87b7c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7e0364\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a4917f\": 1.0, \"candidate_d87b7c\": 0.0, \"candidate_7e0364\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:fba459d327f90e5e3809ddbd:choice", "group_id": "fresh12:snli:fba459d327f90e5e3809ddbd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A large group of protesters are walking down the street with signs.\\nHypothesis: Some people are holding up signs of protest in the street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f4ce9d\": \"The hypothesis follows from the premise.\", \"candidate_ff0945\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1f4587\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f4ce9d\": 1.0, \"candidate_ff0945\": 0.0, \"candidate_1f4587\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2219e12288d8b21584072bcf:choice", "group_id": "fresh12:snli:2219e12288d8b21584072bcf", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in a green flowered dress is holding a container of yarn animals.\\nHypothesis: A lady in a dress has a container of yarn toys.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a3d0cd\": \"The hypothesis follows from the premise.\", \"candidate_b6733a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d67d30\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a3d0cd\": 1.0, \"candidate_b6733a\": 0.0, \"candidate_d67d30\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bf3d04bde8ac5d41c9b586f5:choice", "group_id": "fresh12:snli:bf3d04bde8ac5d41c9b586f5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man wielding an electric razor is gleefully shaving away another man's hair.\\nHypothesis: There are multiple people present.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0c70de\": \"The hypothesis follows from the premise.\", \"candidate_610ada\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_761be1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0c70de\": 1.0, \"candidate_610ada\": 0.0, \"candidate_761be1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a5b278670a851985ed73c89e:choice", "group_id": "fresh12:snli:a5b278670a851985ed73c89e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man stands on a dock near several boats, while 2 other men are on one of the boats.\\nHypothesis: A man is standing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_116a64\": \"The hypothesis follows from the premise.\", \"candidate_3013a7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_87d454\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_116a64\": 1.0, \"candidate_3013a7\": 0.0, \"candidate_87d454\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c9556457dad22336e337a9ad:choice", "group_id": "fresh12:snli:c9556457dad22336e337a9ad", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A guy with piercings, tattoos and a mohawk is looking at the camera while lounging in a chair.\\nHypothesis: A man is in a chair.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_82ae02\": \"The hypothesis follows from the premise.\", \"candidate_e2ce3a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5e8f42\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_82ae02\": 1.0, \"candidate_e2ce3a\": 0.0, \"candidate_5e8f42\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:46a98deddae0d02ed30d7c14:choice", "group_id": "fresh12:snli:46a98deddae0d02ed30d7c14", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A cobblestone street lined with houses and some pedestrians milling around.\\nHypothesis: Humans are outside meandering around.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_78c88c\": \"The hypothesis follows from the premise.\", \"candidate_4c3352\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_381c0e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_78c88c\": 1.0, \"candidate_4c3352\": 0.0, \"candidate_381c0e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:258c28a78d112f495b48cdd9:choice", "group_id": "fresh12:snli:258c28a78d112f495b48cdd9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a black t-shirt and jeans in standing on the sidewalk looking at the camera as people walk by.\\nHypothesis: There is a man in black outside on a busy sidewalk.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_575fd1\": \"The hypothesis follows from the premise.\", \"candidate_a24a32\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e474c3\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_575fd1\": 1.0, \"candidate_a24a32\": 0.0, \"candidate_e474c3\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1c56a25bf61f6e984eb4ac38:choice", "group_id": "fresh12:snli:1c56a25bf61f6e984eb4ac38", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in black walks down the street in front of a graffited wall.\\nHypothesis: A lady walks down the street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_385ac9\": \"The hypothesis follows from the premise.\", \"candidate_3061d0\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9da8fc\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_385ac9\": 1.0, \"candidate_3061d0\": 0.0, \"candidate_9da8fc\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:dbb2a991efd53e5f69f35c77:choice", "group_id": "fresh12:snli:dbb2a991efd53e5f69f35c77", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman wearing bike shorts and a skirt is riding a bike and carrying a shoulder bag.\\nHypothesis: A woman on a bike.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a034a1\": \"The hypothesis follows from the premise.\", \"candidate_4da95d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d69f13\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a034a1\": 1.0, \"candidate_4da95d\": 0.0, \"candidate_d69f13\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5499dcf7171da2fbfa312efa:choice", "group_id": "fresh12:snli:5499dcf7171da2fbfa312efa", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman is lying on her stomach on a white pillow and crying.\\nHypothesis: A woman is crying.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_032784\": \"The hypothesis follows from the premise.\", \"candidate_d9ebfd\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7e6e76\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_032784\": 1.0, \"candidate_d9ebfd\": 0.0, \"candidate_7e6e76\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:65daa26be0aa15d7953ef1c2:choice", "group_id": "fresh12:snli:65daa26be0aa15d7953ef1c2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Boy sitting looking through a glass window at a train station.\\nHypothesis: A boy is sitting at a train station.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5cace2\": \"The hypothesis follows from the premise.\", \"candidate_15926c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8d93c2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5cace2\": 1.0, \"candidate_15926c\": 0.0, \"candidate_8d93c2\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:8a581eb5bb9cde6eae9daba0:choice", "group_id": "fresh12:snli:8a581eb5bb9cde6eae9daba0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: This is a very large vegetable stand under a white tent.\\nHypothesis: Vegetables are being sold under a white tent.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d2241e\": \"The hypothesis follows from the premise.\", \"candidate_d4bb90\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8456df\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d2241e\": 1.0, \"candidate_d4bb90\": 0.0, \"candidate_8456df\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:56384b9dd65557cd011242f8:choice", "group_id": "fresh12:snli:56384b9dd65557cd011242f8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man wearing a white shirt and a hat, riding his bike, makes an obscene gesture to the photographer\\nHypothesis: The man on a bike was rude to the photographer.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_de7dcc\": \"The hypothesis follows from the premise.\", \"candidate_6b9a36\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f443c9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_de7dcc\": 1.0, \"candidate_6b9a36\": 0.0, \"candidate_f443c9\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a5b2b497889c85b444263ea1:choice", "group_id": "fresh12:snli:a5b2b497889c85b444263ea1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A guy on Rollerblades, sliding across a rail.\\nHypothesis: The rollderblader is using the rail.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bd2eed\": \"The hypothesis follows from the premise.\", \"candidate_4d5791\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2deb3d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bd2eed\": 1.0, \"candidate_4d5791\": 0.0, \"candidate_2deb3d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2aea1333e49318f948b19a85:choice", "group_id": "fresh12:snli:2aea1333e49318f948b19a85", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man of the cloth puts a black substance on a man's forehead.\\nHypothesis: The man puts something on the other mans head.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b9c2d4\": \"The hypothesis follows from the premise.\", \"candidate_67634a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f849ad\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b9c2d4\": 1.0, \"candidate_67634a\": 0.0, \"candidate_f849ad\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ce6fb42c64f780b051804d58:choice", "group_id": "fresh12:snli:ce6fb42c64f780b051804d58", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A little blond boy is playing doctor with red, blue, and yellow toy gear.\\nHypothesis: Boy pretending to be a doctor.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_de3e79\": \"The hypothesis follows from the premise.\", \"candidate_983f83\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2331ce\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_de3e79\": 1.0, \"candidate_983f83\": 0.0, \"candidate_2331ce\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7671c81307d33029bb4870d5:choice", "group_id": "fresh12:snli:7671c81307d33029bb4870d5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two boys in front of a soda machine.\\nHypothesis: Two boys are at a machine.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_026a61\": \"The hypothesis follows from the premise.\", \"candidate_9db4fe\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8f85cc\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_026a61\": 1.0, \"candidate_9db4fe\": 0.0, \"candidate_8f85cc\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:abfceba3d9c3c0a96e3cc1bd:choice", "group_id": "fresh12:snli:abfceba3d9c3c0a96e3cc1bd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: In a chain link enclosed skate park a skater goes over one of the obstacles.Wit.\\nHypothesis: A person is skating.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5ae9f2\": \"The hypothesis follows from the premise.\", \"candidate_92c8eb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_467a13\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5ae9f2\": 1.0, \"candidate_92c8eb\": 0.0, \"candidate_467a13\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a3bd196e68bf809a8cd75cec:choice", "group_id": "fresh12:snli:a3bd196e68bf809a8cd75cec", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in the bucket of a chery-picker cuts down branches using a chainsaw at the end of a long pole.\\nHypothesis: A man is trimming trees.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_354d7a\": \"The hypothesis follows from the premise.\", \"candidate_e7fbd8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_892394\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_354d7a\": 1.0, \"candidate_e7fbd8\": 0.0, \"candidate_892394\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e52ed59241254b800f608042:choice", "group_id": "fresh12:snli:e52ed59241254b800f608042", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a pinstriped suit walks down the street and does n't notice the quirky poster on his left.\\nHypothesis: A male has clothes on and is outside walking down the street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_510dee\": \"The hypothesis follows from the premise.\", \"candidate_0c76db\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f1485b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_510dee\": 1.0, \"candidate_0c76db\": 0.0, \"candidate_f1485b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ecc7dbc1ff4e189a9cae72ba:choice", "group_id": "fresh12:snli:ecc7dbc1ff4e189a9cae72ba", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three children are playing with an illuminated water art display.\\nHypothesis: Three children are interacting with an illuminated water art display.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8fee3c\": \"The hypothesis follows from the premise.\", \"candidate_cc4e96\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c83aa1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8fee3c\": 1.0, \"candidate_cc4e96\": 0.0, \"candidate_c83aa1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bde15aa3ce6e133ae840dc9c:choice", "group_id": "fresh12:snli:bde15aa3ce6e133ae840dc9c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An African person standing in a wheat field.\\nHypothesis: A person standing in a wheat field.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2ff6a1\": \"The hypothesis follows from the premise.\", \"candidate_62297a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d195fa\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2ff6a1\": 1.0, \"candidate_62297a\": 0.0, \"candidate_d195fa\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d1a7fb2409531bcfe257d839:choice", "group_id": "fresh12:snli:d1a7fb2409531bcfe257d839", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man preparing to bat in a baseball game.\\nHypothesis: A man is going to play baseball.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5af5e9\": \"The hypothesis follows from the premise.\", \"candidate_dada50\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_163f3b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5af5e9\": 1.0, \"candidate_dada50\": 0.0, \"candidate_163f3b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:13776c33c92acc6696d95b02:choice", "group_id": "fresh12:snli:13776c33c92acc6696d95b02", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy watches a woman walking amidst many concrete boxes.\\nHypothesis: A boy is watching a woman.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_00796d\": \"The hypothesis follows from the premise.\", \"candidate_d4d71f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_af82ac\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_00796d\": 1.0, \"candidate_d4d71f\": 0.0, \"candidate_af82ac\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2cd6a4acb9dd96fef20c97c4:choice", "group_id": "fresh12:snli:2cd6a4acb9dd96fef20c97c4", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A black man wearing tan shorts, a yellow sweatshirt, and a baseball cap is holding up his hand and talking to an Asian woman posing with another young black man while they are holding red flags with yellow stars.\\nHypothesis: a man in shorts\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a5e56\": \"The hypothesis follows from the premise.\", \"candidate_247d29\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d106ec\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a5e56\": 1.0, \"candidate_247d29\": 0.0, \"candidate_d106ec\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1163b5f92e6d3631825169f6:choice", "group_id": "fresh12:snli:1163b5f92e6d3631825169f6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man dressed in a brown shirt and tan shorts is down on one knee looking at some type of machinery.\\nHypothesis: The man is examining a machine.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b7749a\": \"The hypothesis follows from the premise.\", \"candidate_800a23\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_89f3f4\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b7749a\": 1.0, \"candidate_800a23\": 0.0, \"candidate_89f3f4\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cee0fdc8b6bbe34b47127e61:choice", "group_id": "fresh12:snli:cee0fdc8b6bbe34b47127e61", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A person throws a red toy for a dog to run for.\\nHypothesis: A dog is looking for a toy.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_62f621\": \"The hypothesis follows from the premise.\", \"candidate_990760\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_bed9b1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_62f621\": 1.0, \"candidate_990760\": 0.0, \"candidate_bed9b1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d2ad5a6afdedf0549f80a6bc:choice", "group_id": "fresh12:snli:d2ad5a6afdedf0549f80a6bc", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young man wearing khaki pants and a checked shirt is jumping over large rocks with green mountains and a lagoon in the background.\\nHypothesis: A man is jumping over some rocks near the mountains.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_40253d\": \"The hypothesis follows from the premise.\", \"candidate_2e67e4\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_187458\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_40253d\": 1.0, \"candidate_2e67e4\": 0.0, \"candidate_187458\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a1b0fddf0c98707cbaa81675:choice", "group_id": "fresh12:snli:a1b0fddf0c98707cbaa81675", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in orange sunglasses is playing the drums while speaking into a microphone in front of a large sign.\\nHypothesis: A woman is playing drums and speaking at the same time.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_790d47\": \"The hypothesis follows from the premise.\", \"candidate_b2ed0e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_06de3b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_790d47\": 1.0, \"candidate_b2ed0e\": 0.0, \"candidate_06de3b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:70b3004ded06de57c74fc99f:choice", "group_id": "fresh12:snli:70b3004ded06de57c74fc99f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman carrying a large package on her head walks in front of a building plastered with Portuguese posters advertising a concert.\\nHypothesis: A woman walks outdoors.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7db598\": \"The hypothesis follows from the premise.\", \"candidate_f08189\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c38be8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7db598\": 1.0, \"candidate_f08189\": 0.0, \"candidate_c38be8\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e9b1fefae1916f9e8c506bae:choice", "group_id": "fresh12:snli:e9b1fefae1916f9e8c506bae", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man and a boy behind the wheel of a car.\\nHypothesis: The little boy and the man sit behind the car's steering wheel.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6c15f7\": \"The hypothesis follows from the premise.\", \"candidate_8299f8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_426e4c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6c15f7\": 1.0, \"candidate_8299f8\": 0.0, \"candidate_426e4c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:942684043768f2e23a672d22:choice", "group_id": "fresh12:snli:942684043768f2e23a672d22", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Women are working and digging in a field.\\nHypothesis: Women are digging.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_42f23c\": \"The hypothesis follows from the premise.\", \"candidate_f6cc71\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3546db\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_42f23c\": 1.0, \"candidate_f6cc71\": 0.0, \"candidate_3546db\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b824f35688770f0ae1704d64:choice", "group_id": "fresh12:snli:b824f35688770f0ae1704d64", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A little girl wearing a yellow dress is sitting on the white rocks.\\nHypothesis: A little girl is wearing a dress.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a7fbe4\": \"The hypothesis follows from the premise.\", \"candidate_02214e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2461b2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a7fbe4\": 1.0, \"candidate_02214e\": 0.0, \"candidate_2461b2\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:dfb4730b9af4e3792929e19d:choice", "group_id": "fresh12:snli:dfb4730b9af4e3792929e19d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An older woman is holding a small child.\\nHypothesis: a woman holding a child\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6ca108\": \"The hypothesis follows from the premise.\", \"candidate_30e5a1\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a2bb1d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6ca108\": 1.0, \"candidate_30e5a1\": 0.0, \"candidate_a2bb1d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:787ad221277c4675285b07a0:choice", "group_id": "fresh12:snli:787ad221277c4675285b07a0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of people sitting around a table eating food.\\nHypothesis: People are at a table eating food.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fc2fe2\": \"The hypothesis follows from the premise.\", \"candidate_f1fe19\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_22154c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fc2fe2\": 1.0, \"candidate_f1fe19\": 0.0, \"candidate_22154c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c5ae45c2ccf333f0f87868e6:choice", "group_id": "fresh12:snli:c5ae45c2ccf333f0f87868e6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman, standing behind a girl, helping the girl with an experiment.\\nHypothesis: A woman is helping a girl.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b46a8d\": \"The hypothesis follows from the premise.\", \"candidate_653d90\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_709494\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b46a8d\": 1.0, \"candidate_653d90\": 0.0, \"candidate_709494\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5a7fbbcfd84deff2d087e2fb:choice", "group_id": "fresh12:snli:5a7fbbcfd84deff2d087e2fb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man shoveling snow off the roof of a building.\\nHypothesis: A man working in the snow.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0c274e\": \"The hypothesis follows from the premise.\", \"candidate_27a589\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_dea7d9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0c274e\": 1.0, \"candidate_27a589\": 0.0, \"candidate_dea7d9\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1e25bb4e8556a280b58b1842:choice", "group_id": "fresh12:snli:1e25bb4e8556a280b58b1842", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man wearing a black jacket stands next to a body of water.\\nHypothesis: A person is outside near some water.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_679cb6\": \"The hypothesis follows from the premise.\", \"candidate_1c115b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4c0332\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_679cb6\": 1.0, \"candidate_1c115b\": 0.0, \"candidate_4c0332\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:8a5a2f8aa2e5f5065089b32a:choice", "group_id": "fresh12:snli:8a5a2f8aa2e5f5065089b32a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Man in white sweater looks from his press box onto the pitch.\\nHypothesis: A man watches the field\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_01b6cd\": \"The hypothesis follows from the premise.\", \"candidate_fc459c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_648c36\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_01b6cd\": 1.0, \"candidate_fc459c\": 0.0, \"candidate_648c36\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7135dcaaba3e21560ccb5f25:choice", "group_id": "fresh12:snli:7135dcaaba3e21560ccb5f25", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Man with an orange bandanna is playing the flute.\\nHypothesis: Person with an orange bandanna is playing the flute.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e601ad\": \"The hypothesis follows from the premise.\", \"candidate_5da851\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7ec682\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e601ad\": 1.0, \"candidate_5da851\": 0.0, \"candidate_7ec682\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f9a807505da7fd4b97f48a9d:choice", "group_id": "fresh12:snli:f9a807505da7fd4b97f48a9d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The dog is in the snow in front of some trees.\\nHypothesis: A dog is in the snow.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_107d68\": \"The hypothesis follows from the premise.\", \"candidate_8a1595\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f385db\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_107d68\": 1.0, \"candidate_8a1595\": 0.0, \"candidate_f385db\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:33d70706e82eb32f02b74d0c:choice", "group_id": "fresh12:snli:33d70706e82eb32f02b74d0c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An elderly lady holding a scruffy dog and smiling contently.\\nHypothesis: a woman holding her scruffy dog\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_772145\": \"The hypothesis follows from the premise.\", \"candidate_fdb6fa\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e021aa\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_772145\": 1.0, \"candidate_fdb6fa\": 0.0, \"candidate_e021aa\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5054263c742ccffe31ed609e:choice", "group_id": "fresh12:snli:5054263c742ccffe31ed609e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man with a beard is playing an acoustic guitar on the street.\\nHypothesis: A man plays guitar outside.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a14f5\": \"The hypothesis follows from the premise.\", \"candidate_aae81f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e2165f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a14f5\": 1.0, \"candidate_aae81f\": 0.0, \"candidate_e2165f\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1e21d2c7af56e53f7955abc1:choice", "group_id": "fresh12:snli:1e21d2c7af56e53f7955abc1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A female softball player wearing blue and red crouches in the infield, waiting for the next play.\\nHypothesis: a player is waiting\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5f0132\": \"The hypothesis follows from the premise.\", \"candidate_917539\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_39c3a5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5f0132\": 1.0, \"candidate_917539\": 0.0, \"candidate_39c3a5\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c6f6f3cf57aad42d85555a9e:choice", "group_id": "fresh12:snli:c6f6f3cf57aad42d85555a9e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The squirrel sits straight up in the grass.\\nHypothesis: A squirrel sits in the grass.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5caf39\": \"The hypothesis follows from the premise.\", \"candidate_97c4e1\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_227e8d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5caf39\": 1.0, \"candidate_97c4e1\": 0.0, \"candidate_227e8d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:67077247fd9a9a0badf88c1d:choice", "group_id": "fresh12:snli:67077247fd9a9a0badf88c1d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A person rolls down a hill riding a wagon as another watches.\\nHypothesis: Two people are outside.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_138393\": \"The hypothesis follows from the premise.\", \"candidate_b554a2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_06b09e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_138393\": 1.0, \"candidate_b554a2\": 0.0, \"candidate_06b09e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9647201401e72b5b2877e9bd:choice", "group_id": "fresh12:snli:9647201401e72b5b2877e9bd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young boy in green practices juggling in a parking lot.\\nHypothesis: The young boy is outside.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1238da\": \"The hypothesis follows from the premise.\", \"candidate_f2c415\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1af90b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1238da\": 1.0, \"candidate_f2c415\": 0.0, \"candidate_1af90b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6b0f4a90b84e5d44fedff6d6:choice", "group_id": "fresh12:snli:6b0f4a90b84e5d44fedff6d6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A bald person in green clothing stands in front of a brick wall.\\nHypothesis: A person is in front of a brick wall.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_977cab\": \"The hypothesis follows from the premise.\", \"candidate_5f0fed\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1d66e4\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_977cab\": 1.0, \"candidate_5f0fed\": 0.0, \"candidate_1d66e4\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0c542a135c2dbceaddf34c8b:choice", "group_id": "fresh12:snli:0c542a135c2dbceaddf34c8b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a young man pouring tea inside of a glass jar.\\nHypothesis: A man is pouring a glass of tea.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_36495a\": \"The hypothesis follows from the premise.\", \"candidate_799d08\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5a4026\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_36495a\": 1.0, \"candidate_799d08\": 0.0, \"candidate_5a4026\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:08c23b6e8f41b2b987ecf362:choice", "group_id": "fresh12:snli:08c23b6e8f41b2b987ecf362", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man sitting on steps with a child as a bystander walks by.\\nHypothesis: A man and a child sitting on steps saw a woman walk by.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_41d5c3\": \"The hypothesis follows from the premise.\", \"candidate_50b5bd\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d65af2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_41d5c3\": 1.0, \"candidate_50b5bd\": 0.0, \"candidate_d65af2\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4298a1a160b19281cac8d237:choice", "group_id": "fresh12:snli:4298a1a160b19281cac8d237", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A photographer is taking photos of a hockey game, but he is protecting himself from harm by wearing a helmet.\\nHypothesis: A photographer is taking photos of a hockey game\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_59b3c1\": \"The hypothesis follows from the premise.\", \"candidate_1c9036\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_bab433\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_59b3c1\": 1.0, \"candidate_1c9036\": 0.0, \"candidate_bab433\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c983f0963759c541eca83d79:choice", "group_id": "fresh12:snli:c983f0963759c541eca83d79", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a army uniform is holding a picture with trees in it.\\nHypothesis: There is one man in the picture.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0e029c\": \"The hypothesis follows from the premise.\", \"candidate_c0e1ee\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0fc542\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0e029c\": 1.0, \"candidate_c0e1ee\": 0.0, \"candidate_0fc542\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cb8be740afbd5a99ba0e7d50:choice", "group_id": "fresh12:snli:cb8be740afbd5a99ba0e7d50", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two small boys in blue soccer uniforms use a wooden set of steps to wash their hands in an adult-sized bathroom.\\nHypothesis: Two boys washing their hands\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a54d10\": \"The hypothesis follows from the premise.\", \"candidate_65ca81\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_944e45\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a54d10\": 1.0, \"candidate_65ca81\": 0.0, \"candidate_944e45\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f048ca8df5d9a4e4cf1e482a:choice", "group_id": "fresh12:snli:f048ca8df5d9a4e4cf1e482a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A smiling girl slides down a purple slide feet first.\\nHypothesis: A girl goes down a slide.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_16cde9\": \"The hypothesis follows from the premise.\", \"candidate_6c6e62\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_596200\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_16cde9\": 1.0, \"candidate_6c6e62\": 0.0, \"candidate_596200\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4b5c70578a26e681a88e077e:choice", "group_id": "fresh12:snli:4b5c70578a26e681a88e077e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A blond-haired baby is sitting on the floor playing with toys while looking at a black and white cat.\\nHypothesis: The baby and the cat are in the same room\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_50ddd2\": \"The hypothesis follows from the premise.\", \"candidate_e23ebf\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a967e0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_50ddd2\": 1.0, \"candidate_e23ebf\": 0.0, \"candidate_a967e0\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:02ee012732804794bcc36fb6:choice", "group_id": "fresh12:snli:02ee012732804794bcc36fb6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman wearing a red helmet is holding a rope and smiling.\\nHypothesis: A happy woman is holding rope.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0ed40a\": \"The hypothesis follows from the premise.\", \"candidate_34546c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5f56ce\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0ed40a\": 1.0, \"candidate_34546c\": 0.0, \"candidate_5f56ce\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:90574f4a48e8210aa656edab:choice", "group_id": "fresh12:snli:90574f4a48e8210aa656edab", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two black labs run, one carrying an orange dog toy.\\nHypothesis: two dogs are running.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a1e594\": \"The hypothesis follows from the premise.\", \"candidate_a58d54\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_89c4f1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a1e594\": 1.0, \"candidate_a58d54\": 0.0, \"candidate_89c4f1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e10d23e9a9828ecd6b4d9ce2:choice", "group_id": "fresh12:snli:e10d23e9a9828ecd6b4d9ce2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Bubbles surround a statue in the middle of a street.\\nHypothesis: There are bubbles around the statue.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_064873\": \"The hypothesis follows from the premise.\", \"candidate_e0e627\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a8b9a5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_064873\": 1.0, \"candidate_e0e627\": 0.0, \"candidate_a8b9a5\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:38e86ecfe8ff399fcd39e2b7:choice", "group_id": "fresh12:snli:38e86ecfe8ff399fcd39e2b7", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man is holding a child on his shoulders.\\nHypothesis: A child is riding on a mans shuolders.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_75a785\": \"The hypothesis follows from the premise.\", \"candidate_c32559\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ab0760\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_75a785\": 1.0, \"candidate_c32559\": 0.0, \"candidate_ab0760\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7f3b08db08350a05758d6152:choice", "group_id": "fresh12:snli:7f3b08db08350a05758d6152", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A basketball player in a red uniform is trying to make a basket but is being blocked by the opposing team.\\nHypothesis: A basketball player is playing basketball.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4ba13e\": \"The hypothesis follows from the premise.\", \"candidate_a01029\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c04be1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4ba13e\": 1.0, \"candidate_a01029\": 0.0, \"candidate_c04be1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c6048e0a17820b1140090561:choice", "group_id": "fresh12:snli:c6048e0a17820b1140090561", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Casually dressed young males and females are sitting and laying on a grassy area with bicycles and other personal belongings reading on a sunny day.\\nHypothesis: people are hanging outside with their bicycles on the grass.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3721a5\": \"The hypothesis follows from the premise.\", \"candidate_acce1f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a4c394\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3721a5\": 1.0, \"candidate_acce1f\": 0.0, \"candidate_a4c394\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d8272382c11cf4f4d561a63b:choice", "group_id": "fresh12:snli:d8272382c11cf4f4d561a63b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy lying on a hospital bed with his leg hanging over the side.\\nHypothesis: a kid is at the hospital\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_10d7e7\": \"The hypothesis follows from the premise.\", \"candidate_cd279d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ba55e2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_10d7e7\": 1.0, \"candidate_cd279d\": 0.0, \"candidate_ba55e2\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f91c1b7d2074cd4a0c3105b3:choice", "group_id": "fresh12:snli:f91c1b7d2074cd4a0c3105b3", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An impoverished person, wearing a torn brown shirt and no shoes, is kneeling down along the banks of a trash infested body of water.\\nHypothesis: A poor person in tattered clothes kneels by the trash filled body of water.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7fb1ba\": \"The hypothesis follows from the premise.\", \"candidate_b27660\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_309550\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7fb1ba\": 1.0, \"candidate_b27660\": 0.0, \"candidate_309550\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2ba726d9570ab084a91a2233:choice", "group_id": "fresh12:snli:2ba726d9570ab084a91a2233", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young female gymnast, wearing a sequined royal blue and white outfit, is lifting her right leg straight up in the air.\\nHypothesis: a gymnast has her leg straight up in the air.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_34c55c\": \"The hypothesis follows from the premise.\", \"candidate_5cefbe\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c1dee8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_34c55c\": 1.0, \"candidate_5cefbe\": 0.0, \"candidate_c1dee8\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:17ae001355b1da86c65a1e03:choice", "group_id": "fresh12:snli:17ae001355b1da86c65a1e03", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A small girl with short brown hair is pushing a pink scooter over grass with brown leaves with threes in the background\\nHypothesis: The girl is outside.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_486cd2\": \"The hypothesis follows from the premise.\", \"candidate_001b75\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_16d2ab\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_486cd2\": 1.0, \"candidate_001b75\": 0.0, \"candidate_16d2ab\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7ce2fb5675b6d086b6238437:choice", "group_id": "fresh12:snli:7ce2fb5675b6d086b6238437", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A person with a striped tube standing in the middle of a crowd of people.\\nHypothesis: A human is surrounded.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b0fc82\": \"The hypothesis follows from the premise.\", \"candidate_a64fb4\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7ca4d9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b0fc82\": 1.0, \"candidate_a64fb4\": 0.0, \"candidate_7ca4d9\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:08a803e6cf0a5c6a97793f03:choice", "group_id": "fresh12:snli:08a803e6cf0a5c6a97793f03", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: One young boy in a yellow long-sleeved shirt and jeans sitting down, with five adults in front of him, one operating a video camera.\\nHypothesis: An adult using a video camera.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c58acc\": \"The hypothesis follows from the premise.\", \"candidate_6f9579\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f8c749\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c58acc\": 1.0, \"candidate_6f9579\": 0.0, \"candidate_f8c749\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:84af2ad35f93ede4ab6ae068:choice", "group_id": "fresh12:snli:84af2ad35f93ede4ab6ae068", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A girl jumping high off of the sand.\\nHypothesis: There is a girl on the sand.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ceb01f\": \"The hypothesis follows from the premise.\", \"candidate_c85516\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_57b9c0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ceb01f\": 1.0, \"candidate_c85516\": 0.0, \"candidate_57b9c0\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9687fcd687eff06f7dccd24e:choice", "group_id": "fresh12:snli:9687fcd687eff06f7dccd24e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in a yellow skirt is dancing with a man in a white shirt, with another couple dancing in the background.\\nHypothesis: A woman dances with a man next to another couple dancing\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2026a0\": \"The hypothesis follows from the premise.\", \"candidate_177801\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_fa4384\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2026a0\": 1.0, \"candidate_177801\": 0.0, \"candidate_fa4384\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6f8167ac0c2cc2ec5184684f:choice", "group_id": "fresh12:snli:6f8167ac0c2cc2ec5184684f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three men are doing a musical set with wires and equipment all over the stage, and they are holding unique guitar like instruments.\\nHypothesis: Three men are doing a musical set wit wires and equipment all over the stage and they are holding unique guitar like instruments\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ac1632\": \"The hypothesis follows from the premise.\", \"candidate_82a77a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b9331c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ac1632\": 1.0, \"candidate_82a77a\": 0.0, \"candidate_b9331c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:75cd9642f036f5def1f68f19:choice", "group_id": "fresh12:snli:75cd9642f036f5def1f68f19", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A topless man straps surfboards on top of his blue car.\\nHypothesis: A man secures surfboards on his car.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7ea8e7\": \"The hypothesis follows from the premise.\", \"candidate_653b0d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7b377d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7ea8e7\": 1.0, \"candidate_653b0d\": 0.0, \"candidate_7b377d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2215e357bd37817719a771a0:choice", "group_id": "fresh12:snli:2215e357bd37817719a771a0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in red shows is bowling with a red ball at a bowling lane with Tecate ads.\\nHypothesis: The man is bowling.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a57377\": \"The hypothesis follows from the premise.\", \"candidate_064498\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5230e2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a57377\": 1.0, \"candidate_064498\": 0.0, \"candidate_5230e2\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:37d676ed4162a6697ffbe421:choice", "group_id": "fresh12:snli:37d676ed4162a6697ffbe421", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Several people including a child and a clown are walking towards a snowy sidewalk\\nHypothesis: Many people walk to the sidewalk.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_19dae9\": \"The hypothesis follows from the premise.\", \"candidate_791615\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5d233d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_19dae9\": 1.0, \"candidate_791615\": 0.0, \"candidate_5d233d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:646883a8306501247cd5c7a6:choice", "group_id": "fresh12:snli:646883a8306501247cd5c7a6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Four men playing drums in very orange lighting while one of them is also drinking something out of a bottle.\\nHypothesis: four men play drums\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bfde18\": \"The hypothesis follows from the premise.\", \"candidate_9201ab\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_815392\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bfde18\": 1.0, \"candidate_9201ab\": 0.0, \"candidate_815392\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4379cdbc885388b80423642a:choice", "group_id": "fresh12:snli:4379cdbc885388b80423642a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man is making a phone call.\\nHypothesis: A man is on the phone.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3f3f89\": \"The hypothesis follows from the premise.\", \"candidate_9961d9\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e8e62a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3f3f89\": 1.0, \"candidate_9961d9\": 0.0, \"candidate_e8e62a\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:053592938550a150d21d31dc:choice", "group_id": "fresh12:snli:053592938550a150d21d31dc", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in black walks past a shop selling a t-shirt depicting a tattooed woman smoking a cigarette.\\nHypothesis: A woman is walking down the street of shops.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e4c5e6\": \"The hypothesis follows from the premise.\", \"candidate_70fba7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_53cc64\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e4c5e6\": 1.0, \"candidate_70fba7\": 0.0, \"candidate_53cc64\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:37683b861867f13eccd3587b:choice", "group_id": "fresh12:snli:37683b861867f13eccd3587b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in red blouse is standing with small blond child in front of a small folding chalkboard.\\nHypothesis: a woman stands with a child\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f56c90\": \"The hypothesis follows from the premise.\", \"candidate_e61774\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_063dc0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f56c90\": 1.0, \"candidate_e61774\": 0.0, \"candidate_063dc0\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:20262a1958c87d6be18c962a:choice", "group_id": "fresh12:snli:20262a1958c87d6be18c962a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young man dressed up with a yellow tie jumps against a wall and one foot is left against the wall to push away.\\nHypothesis: The young man is jumping against a wall.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c752a4\": \"The hypothesis follows from the premise.\", \"candidate_085133\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e3d3fd\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c752a4\": 1.0, \"candidate_085133\": 0.0, \"candidate_e3d3fd\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e0a3763e9a0518af2df0ff26:choice", "group_id": "fresh12:snli:e0a3763e9a0518af2df0ff26", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A brown and black dog rubs his face in the grass.\\nHypothesis: The dog is outside.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b08a2a\": \"The hypothesis follows from the premise.\", \"candidate_54984e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d786ac\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b08a2a\": 1.0, \"candidate_54984e\": 0.0, \"candidate_d786ac\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6cc8241f3a7a3691d7cc80e6:choice", "group_id": "fresh12:snli:6cc8241f3a7a3691d7cc80e6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young girl, is mid jump, while performing a dance routine.\\nHypothesis: A young girl is dancing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c0ab71\": \"The hypothesis follows from the premise.\", \"candidate_6c5629\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_063ccc\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c0ab71\": 1.0, \"candidate_6c5629\": 0.0, \"candidate_063ccc\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:8b30d3d3dd05ae9546428f44:choice", "group_id": "fresh12:snli:8b30d3d3dd05ae9546428f44", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Six people appearing to be planting something in the ground.\\nHypothesis: People are digging up the earth\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eee65b\": \"The hypothesis follows from the premise.\", \"candidate_31bd40\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_348a41\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eee65b\": 1.0, \"candidate_31bd40\": 0.0, \"candidate_348a41\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:757449d73a59be3481dfe6f3:choice", "group_id": "fresh12:snli:757449d73a59be3481dfe6f3", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young girl is playing a musical instrument and singing into a microphone.\\nHypothesis: A talented child sings into a mic and plays an instrument.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5c4b4d\": \"The hypothesis follows from the premise.\", \"candidate_0a0d54\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d3c509\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5c4b4d\": 1.0, \"candidate_0a0d54\": 0.0, \"candidate_d3c509\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:572f95ddee648b3d74cacdce:choice", "group_id": "fresh12:snli:572f95ddee648b3d74cacdce", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A band playing on stage in a club.\\nHypothesis: Musicians are playing their instruments at a venue\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_404109\": \"The hypothesis follows from the premise.\", \"candidate_233128\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3670c7\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_404109\": 1.0, \"candidate_233128\": 0.0, \"candidate_3670c7\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a9ebc83bf413c048092161b9:choice", "group_id": "fresh12:snli:a9ebc83bf413c048092161b9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young man in a heavy brown winter coat stands in front of a blue railing with his arms spread.\\nHypothesis: The man spreads his arms while standing in front of a railing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a495c3\": \"The hypothesis follows from the premise.\", \"candidate_9e0f74\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a5fa1c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a495c3\": 1.0, \"candidate_9e0f74\": 0.0, \"candidate_a5fa1c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:3611758179b884350f17b227:choice", "group_id": "fresh12:snli:3611758179b884350f17b227", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: This is when daddy taught her how to ride her scooter.\\nHypothesis: daddy taught her\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ed1b60\": \"The hypothesis follows from the premise.\", \"candidate_e485af\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_186be4\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ed1b60\": 1.0, \"candidate_e485af\": 0.0, \"candidate_186be4\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1350916af89bdc483bdda927:choice", "group_id": "fresh12:snli:1350916af89bdc483bdda927", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A female swimmer wearing a swimming cap does the butterfly stroke.\\nHypothesis: A woman is swimming.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_888280\": \"The hypothesis follows from the premise.\", \"candidate_91d77c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f4d922\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_888280\": 1.0, \"candidate_91d77c\": 0.0, \"candidate_f4d922\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:73858b49ed9ff2d2440e85ff:choice", "group_id": "fresh12:snli:73858b49ed9ff2d2440e85ff", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A few young boys throwing a football over a volleyball net.\\nHypothesis: Boys playing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f4c9ba\": \"The hypothesis follows from the premise.\", \"candidate_bffc1c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7d99ae\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f4c9ba\": 1.0, \"candidate_bffc1c\": 0.0, \"candidate_7d99ae\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:69729a0f533a0ba32d68459f:choice", "group_id": "fresh12:snli:69729a0f533a0ba32d68459f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Several women are playing volleyball.\\nHypothesis: They women are hitting a ball with their arms\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_668291\": \"The hypothesis follows from the premise.\", \"candidate_9a18d2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_88f96d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_668291\": 1.0, \"candidate_9a18d2\": 0.0, \"candidate_88f96d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:21ab38b1bd4d3184608ca249:choice", "group_id": "fresh12:snli:21ab38b1bd4d3184608ca249", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: two dogs run across the woods.\\nHypothesis: The canines run around in the forest.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_86a506\": \"The hypothesis follows from the premise.\", \"candidate_a7c952\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7d10aa\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_86a506\": 1.0, \"candidate_a7c952\": 0.0, \"candidate_7d10aa\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1d911faeb0ac92af88ea3f07:choice", "group_id": "fresh12:snli:1d911faeb0ac92af88ea3f07", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A dad with his child and an apple pie.\\nHypothesis: A dad is with his child.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0decfa\": \"The hypothesis follows from the premise.\", \"candidate_123d64\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_942daa\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0decfa\": 1.0, \"candidate_123d64\": 0.0, \"candidate_942daa\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0c4aad31e9e578f89b1dee8f:choice", "group_id": "fresh12:snli:0c4aad31e9e578f89b1dee8f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The man in white is playing basketball against the man in blue.\\nHypothesis: Two men are engaging in a sporting activity.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ee1507\": \"The hypothesis follows from the premise.\", \"candidate_736246\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8ce0e7\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ee1507\": 1.0, \"candidate_736246\": 0.0, \"candidate_8ce0e7\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:dbd48f0e642244a86f7230bb:choice", "group_id": "fresh12:snli:dbd48f0e642244a86f7230bb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman player in white and black, sends the volleyball over the net, while the opposing team in navy blue, tries to block it, as the coach and a few spectators watch, in a gym.\\nHypothesis: They are in a gym\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_495c04\": \"The hypothesis follows from the premise.\", \"candidate_6786dc\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_cb86c0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_495c04\": 1.0, \"candidate_6786dc\": 0.0, \"candidate_cb86c0\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7a1ac4ab45493e2490c4cc8c:choice", "group_id": "fresh12:snli:7a1ac4ab45493e2490c4cc8c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The indian's are performing a sacred dance.\\nHypothesis: The people are dancing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f5b755\": \"The hypothesis follows from the premise.\", \"candidate_3b1e4b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b3af65\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f5b755\": 1.0, \"candidate_3b1e4b\": 0.0, \"candidate_b3af65\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2cbc6f0ebc26ea3ce71dbbd5:choice", "group_id": "fresh12:snli:2cbc6f0ebc26ea3ce71dbbd5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in renaissance clothing is sitting on a brick wall.\\nHypothesis: The man is in costume\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_828705\": \"The hypothesis follows from the premise.\", \"candidate_2f03df\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a07c96\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_828705\": 1.0, \"candidate_2f03df\": 0.0, \"candidate_a07c96\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6e3886d6bfe9c4341049a0a8:choice", "group_id": "fresh12:snli:6e3886d6bfe9c4341049a0a8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A couple stands at the top of a flight of outdoor stairs kissing one another.\\nHypothesis: A couple stands near some stairs.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8d1ba6\": \"The hypothesis follows from the premise.\", \"candidate_5c8974\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3ddafb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8d1ba6\": 1.0, \"candidate_5c8974\": 0.0, \"candidate_3ddafb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:150520cd8cfc57f1b65d9518:choice", "group_id": "fresh12:snli:150520cd8cfc57f1b65d9518", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: People shopping at an outside market\\nHypothesis: People are outside.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a2fcb\": \"The hypothesis follows from the premise.\", \"candidate_99b441\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_baec57\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a2fcb\": 1.0, \"candidate_99b441\": 0.0, \"candidate_baec57\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5c14756c1292c6c3099f0320:choice", "group_id": "fresh12:snli:5c14756c1292c6c3099f0320", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man with a pinstriped shirt and tan slacks adjust his tie while looking at a apple laptop computer.\\nHypothesis: A man adjusts his new tie in front of a laptop\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_618400\": \"The hypothesis follows from the premise.\", \"candidate_883a01\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_03b653\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_618400\": 1.0, \"candidate_883a01\": 0.0, \"candidate_03b653\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e30be61dbf89abcaa874ac82:choice", "group_id": "fresh12:snli:e30be61dbf89abcaa874ac82", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Times Square in New York.\\nHypothesis: Times Square is a place.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bd4646\": \"The hypothesis follows from the premise.\", \"candidate_24108f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0e1927\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bd4646\": 1.0, \"candidate_24108f\": 0.0, \"candidate_0e1927\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c5ee1b41408286d410561043:choice", "group_id": "fresh12:snli:c5ee1b41408286d410561043", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three boys in white shirts are walking behind an older model Ford car.\\nHypothesis: The boys are on the street\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c38e48\": \"The hypothesis follows from the premise.\", \"candidate_7a8f21\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_79596d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c38e48\": 1.0, \"candidate_7a8f21\": 0.0, \"candidate_79596d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5e2eb1f7365b71de6e33798c:choice", "group_id": "fresh12:snli:5e2eb1f7365b71de6e33798c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young boy vocalizes in to a microphone held by a blond woman in black.\\nHypothesis: A boy singing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4687a5\": \"The hypothesis follows from the premise.\", \"candidate_002429\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_6af010\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4687a5\": 1.0, \"candidate_002429\": 0.0, \"candidate_6af010\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c31f04a84a94755b066c8aa8:choice", "group_id": "fresh12:snli:c31f04a84a94755b066c8aa8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two beautiful, little girls with dark hair and dark complexions are sitting down eating cupcakes.\\nHypothesis: Two girls eating cupcakes.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_af1156\": \"The hypothesis follows from the premise.\", \"candidate_b12310\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_549581\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_af1156\": 1.0, \"candidate_b12310\": 0.0, \"candidate_549581\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0183a290220e7a161e2655d4:choice", "group_id": "fresh12:snli:0183a290220e7a161e2655d4", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two women drink coffee and chat while sitting on a blanket near a rocky shore.\\nHypothesis: Both women are drinking coffee outside.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a97432\": \"The hypothesis follows from the premise.\", \"candidate_0c5462\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_830c33\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a97432\": 1.0, \"candidate_0c5462\": 0.0, \"candidate_830c33\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7deb8403e69f566014e09907:choice", "group_id": "fresh12:snli:7deb8403e69f566014e09907", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A crowd gathers and watches a bonfire.\\nHypothesis: A crowd sees flames.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bbea24\": \"The hypothesis follows from the premise.\", \"candidate_0631b9\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_afcff8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bbea24\": 1.0, \"candidate_0631b9\": 0.0, \"candidate_afcff8\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:042e24e75c40faff571b775d:choice", "group_id": "fresh12:snli:042e24e75c40faff571b775d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A Young girl wearing a blue ski cap and red and gray two tone jacket stands between two bits of ice.\\nHypothesis: A girl is standing in a winter landscape.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a472eb\": \"The hypothesis follows from the premise.\", \"candidate_03d832\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_738ac1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a472eb\": 1.0, \"candidate_03d832\": 0.0, \"candidate_738ac1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a8e0539fc5dff20d7c091ea5:choice", "group_id": "fresh12:snli:a8e0539fc5dff20d7c091ea5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A little girl on a piece of playground equipment\\nHypothesis: A kid is doing a thing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_11618e\": \"The hypothesis follows from the premise.\", \"candidate_897cd0\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_100071\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_11618e\": 1.0, \"candidate_897cd0\": 0.0, \"candidate_100071\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d0f3226b5d04d338c352cd45:choice", "group_id": "fresh12:snli:d0f3226b5d04d338c352cd45", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man wearing shorts looks at his cellphone while a woman walks outside a small building.\\nHypothesis: There are multiple people present.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bbd36a\": \"The hypothesis follows from the premise.\", \"candidate_37ef96\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2f5112\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bbd36a\": 1.0, \"candidate_37ef96\": 0.0, \"candidate_2f5112\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7a8c71b6490eaf4055b5a275:choice", "group_id": "fresh12:snli:7a8c71b6490eaf4055b5a275", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: 3 ladies with numbers on their shirts running.\\nHypothesis: There are ladies running.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2c03c4\": \"The hypothesis follows from the premise.\", \"candidate_324f96\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3a5304\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2c03c4\": 1.0, \"candidate_324f96\": 0.0, \"candidate_3a5304\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1e1a05f681ad2f9f28b10f33:choice", "group_id": "fresh12:snli:1e1a05f681ad2f9f28b10f33", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two roadside workers with lime green safety jackets, white hard hats and gloves on with construction cones in the background\\nHypothesis: Two men are wearing lime green jackets.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_62c6ce\": \"The hypothesis follows from the premise.\", \"candidate_4caea1\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d13ddb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_62c6ce\": 1.0, \"candidate_4caea1\": 0.0, \"candidate_d13ddb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bd1d2e5da338be46c5a815f4:choice", "group_id": "fresh12:snli:bd1d2e5da338be46c5a815f4", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An accident involving a green Sudan and a gray van.\\nHypothesis: Two cars were in an accident.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ee4542\": \"The hypothesis follows from the premise.\", \"candidate_6fb695\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_53b013\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ee4542\": 1.0, \"candidate_6fb695\": 0.0, \"candidate_53b013\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:83b5412c83ee5c80af9551b2:choice", "group_id": "fresh12:snli:83b5412c83ee5c80af9551b2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young blond child eats chocolate cake with his fingers while holding the spoon in their other hand.\\nHypothesis: a child eats food with his hands\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_df179c\": \"The hypothesis follows from the premise.\", \"candidate_e487ff\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_515bc0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_df179c\": 1.0, \"candidate_e487ff\": 0.0, \"candidate_515bc0\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9081ebc4c62dda27ebd31e74:choice", "group_id": "fresh12:snli:9081ebc4c62dda27ebd31e74", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A worker peers out from atop a building under construction.\\nHypothesis: A person is atop of a building.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f1d56c\": \"The hypothesis follows from the premise.\", \"candidate_40eb5c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_070175\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f1d56c\": 1.0, \"candidate_40eb5c\": 0.0, \"candidate_070175\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d63656a5af49a9e73546fb26:choice", "group_id": "fresh12:snli:d63656a5af49a9e73546fb26", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A male with sunglasses is poking at a tree with a pole.\\nHypothesis: A human poking\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1a7f9d\": \"The hypothesis follows from the premise.\", \"candidate_c3448a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_165b91\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1a7f9d\": 1.0, \"candidate_c3448a\": 0.0, \"candidate_165b91\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5a0a7d2d891a1ce4d95d4816:choice", "group_id": "fresh12:snli:5a0a7d2d891a1ce4d95d4816", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: One girl sips a soda while another looks on, standing on a street in front of a bunch of bicycles.\\nHypothesis: A girl drinks a soda on the street\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ac60fe\": \"The hypothesis follows from the premise.\", \"candidate_bbbe7d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_702f4c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ac60fe\": 1.0, \"candidate_bbbe7d\": 0.0, \"candidate_702f4c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:40ee6cc35caf585996fc3ec8:choice", "group_id": "fresh12:snli:40ee6cc35caf585996fc3ec8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: this Man, with a Red& White Shirt has Water Bottles on this White Truck.\\nHypothesis: the guy has bottles on the truck\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4d044c\": \"The hypothesis follows from the premise.\", \"candidate_0ed09f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_dfccd7\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4d044c\": 1.0, \"candidate_0ed09f\": 0.0, \"candidate_dfccd7\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5badb6136945483fe1332b87:choice", "group_id": "fresh12:snli:5badb6136945483fe1332b87", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men in suits standing in front of a bus, one is wearing a medal.\\nHypothesis: Two men are standing in front of a large vehicle.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f13bdf\": \"The hypothesis follows from the premise.\", \"candidate_4940e7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_310f54\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f13bdf\": 1.0, \"candidate_4940e7\": 0.0, \"candidate_310f54\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d5f6026964cd0ac34d2b21fd:choice", "group_id": "fresh12:snli:d5f6026964cd0ac34d2b21fd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A blond woman in a black shirt is standing behind a counter.\\nHypothesis: The woman is standing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_09cc09\": \"The hypothesis follows from the premise.\", \"candidate_c689d8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c5d203\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_09cc09\": 1.0, \"candidate_c689d8\": 0.0, \"candidate_c5d203\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:60fad52552e4d10c9c24b096:choice", "group_id": "fresh12:snli:60fad52552e4d10c9c24b096", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Construction worker standing atop an unfinished building.\\nHypothesis: The constructon worker is on the building\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_903057\": \"The hypothesis follows from the premise.\", \"candidate_4173bc\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_29b8ad\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_903057\": 1.0, \"candidate_4173bc\": 0.0, \"candidate_29b8ad\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9db2ae5a6538fd1df7e40df3:choice", "group_id": "fresh12:snli:9db2ae5a6538fd1df7e40df3", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two dogs are digging under a large rock with their bottoms to the camera.\\nHypothesis: Dogs are digging a hole.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4d5974\": \"The hypothesis follows from the premise.\", \"candidate_5cdc26\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_52dbbb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4d5974\": 1.0, \"candidate_5cdc26\": 0.0, \"candidate_52dbbb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:dca7d725e21716895fa59398:choice", "group_id": "fresh12:snli:dca7d725e21716895fa59398", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a little boy is riding a yellow bicycle across a town square.\\nHypothesis: A boy has a yellow bike.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7fc75a\": \"The hypothesis follows from the premise.\", \"candidate_1a22c8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c76837\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7fc75a\": 1.0, \"candidate_1a22c8\": 0.0, \"candidate_c76837\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:aac68cdfce51fd3462d5c104:choice", "group_id": "fresh12:snli:aac68cdfce51fd3462d5c104", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A girl sits at a table in front of a plate of food.\\nHypothesis: The girl is near food.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_40c59c\": \"The hypothesis follows from the premise.\", \"candidate_082bf7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4fe398\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_40c59c\": 1.0, \"candidate_082bf7\": 0.0, \"candidate_4fe398\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:956c65ab4c93e3a761e77859:choice", "group_id": "fresh12:snli:956c65ab4c93e3a761e77859", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: lady in black skirt is getting excited.\\nHypothesis: A lady is excited.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5d61ab\": \"The hypothesis follows from the premise.\", \"candidate_42ab3d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_cd2f08\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5d61ab\": 1.0, \"candidate_42ab3d\": 0.0, \"candidate_cd2f08\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a21f0504d8cab13c7b1687d4:choice", "group_id": "fresh12:snli:a21f0504d8cab13c7b1687d4", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Women in costume stand on a stage.\\nHypothesis: The women are on stage.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8570c7\": \"The hypothesis follows from the premise.\", \"candidate_844d75\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e07bd8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8570c7\": 1.0, \"candidate_844d75\": 0.0, \"candidate_e07bd8\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6c7d2064a6e3fe28774f8f14:choice", "group_id": "fresh12:snli:6c7d2064a6e3fe28774f8f14", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three children in bathing suits are playing in foam\\nHypothesis: some kids play in some foam\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a9f777\": \"The hypothesis follows from the premise.\", \"candidate_674169\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_25831d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a9f777\": 1.0, \"candidate_674169\": 0.0, \"candidate_25831d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:8d91fbfc01bcf98f4df01ffb:choice", "group_id": "fresh12:snli:8d91fbfc01bcf98f4df01ffb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a jacket wearing black boots walking on the truck of a fallen tree over a ravine.\\nHypothesis: A man walks outdoors.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_70c709\": \"The hypothesis follows from the premise.\", \"candidate_d496aa\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f04024\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_70c709\": 1.0, \"candidate_d496aa\": 0.0, \"candidate_f04024\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b40a86a731487d2bfcdd2ebd:choice", "group_id": "fresh12:snli:b40a86a731487d2bfcdd2ebd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two girls stand in front of two men selling what appears to be concert merchandise while a woman holding a microphone interviews the men behind the merchandise counter.\\nHypothesis: girls can stand\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5c330f\": \"The hypothesis follows from the premise.\", \"candidate_c1ebbb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7b0585\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5c330f\": 1.0, \"candidate_c1ebbb\": 0.0, \"candidate_7b0585\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:8a748d2144e5fccf696221c5:choice", "group_id": "fresh12:snli:8a748d2144e5fccf696221c5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young girl standing in front of a row of yellow chairs.\\nHypothesis: A human standing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bdeff5\": \"The hypothesis follows from the premise.\", \"candidate_cbe043\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_776a37\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bdeff5\": 1.0, \"candidate_cbe043\": 0.0, \"candidate_776a37\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cdcc81ee77e34074edc6765e:choice", "group_id": "fresh12:snli:cdcc81ee77e34074edc6765e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two casually-dressed men, one in a cowboy hat and one in a baseball cap, make their way through a crowd.\\nHypothesis: A man is wearing a cowboy hat.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_75f35d\": \"The hypothesis follows from the premise.\", \"candidate_60189e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_63c71b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_75f35d\": 1.0, \"candidate_60189e\": 0.0, \"candidate_63c71b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:39e9087fc4ed5df44cb8ebc7:choice", "group_id": "fresh12:snli:39e9087fc4ed5df44cb8ebc7", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Uniformed group walks down a street.\\nHypothesis: A group of people walk down the street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4219a1\": \"The hypothesis follows from the premise.\", \"candidate_ab1da4\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_51709b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4219a1\": 1.0, \"candidate_ab1da4\": 0.0, \"candidate_51709b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f330ec464d08d59ce67960ad:choice", "group_id": "fresh12:snli:f330ec464d08d59ce67960ad", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A dog attacking a man wearing protective gear.\\nHypothesis: The man protected himself.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3767a1\": \"The hypothesis follows from the premise.\", \"candidate_d85371\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9b6d12\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3767a1\": 1.0, \"candidate_d85371\": 0.0, \"candidate_9b6d12\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:840b84a4b3720f1da821a538:choice", "group_id": "fresh12:snli:840b84a4b3720f1da821a538", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A little boy in flannel pajama pants and an orange shirt is playing the guitar on the couch.\\nHypothesis: The child practices guitar indoors\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_58d390\": \"The hypothesis follows from the premise.\", \"candidate_507a1c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_25628b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_58d390\": 1.0, \"candidate_507a1c\": 0.0, \"candidate_25628b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f493c0ba0c7e1108b2eaff53:choice", "group_id": "fresh12:snli:f493c0ba0c7e1108b2eaff53", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young man sitting under a tree while others walk down the street.\\nHypothesis: A young man is sitting down.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c66d76\": \"The hypothesis follows from the premise.\", \"candidate_96ca40\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d30de4\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c66d76\": 1.0, \"candidate_96ca40\": 0.0, \"candidate_d30de4\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:dd94c8a8edd86ce8a7167cbe:choice", "group_id": "fresh12:snli:dd94c8a8edd86ce8a7167cbe", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A blond woman in a white shirt demonstrates her talents to a crowd.\\nHypothesis: The woman is a street performer.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_95438d\": \"The hypothesis follows from the premise.\", \"candidate_d9eb8e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2644eb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_95438d\": 1.0, \"candidate_d9eb8e\": 0.0, \"candidate_2644eb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:602980a23527c14aa530c2a0:choice", "group_id": "fresh12:snli:602980a23527c14aa530c2a0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A golden dog bounds across the snow-covered hill.\\nHypothesis: A golden dog bounds across the snow-covered land form.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_538a72\": \"The hypothesis follows from the premise.\", \"candidate_aed8f0\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_493f9f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_538a72\": 1.0, \"candidate_aed8f0\": 0.0, \"candidate_493f9f\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:29a018e1c759c84c847b3171:choice", "group_id": "fresh12:snli:29a018e1c759c84c847b3171", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A black man in a white uniform makes a spectacular reverse slam dunk to the crowd's amazement.\\nHypothesis: the man is playing\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0b6479\": \"The hypothesis follows from the premise.\", \"candidate_31a75b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1f9a50\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0b6479\": 1.0, \"candidate_31a75b\": 0.0, \"candidate_1f9a50\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bb5b746a1f265e71a6af7ecb:choice", "group_id": "fresh12:snli:bb5b746a1f265e71a6af7ecb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two little girls, wearing pink hanging on a tube.\\nHypothesis: Two girls playing on a tube.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_077ae7\": \"The hypothesis follows from the premise.\", \"candidate_91f0a2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_246803\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_077ae7\": 1.0, \"candidate_91f0a2\": 0.0, \"candidate_246803\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0b0d85f1c3d428c644859f37:choice", "group_id": "fresh12:snli:0b0d85f1c3d428c644859f37", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man looks at an iPad on a subway, while a beautiful women reads a book.\\nHypothesis: Two people are occupied with different content.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_197a40\": \"The hypothesis follows from the premise.\", \"candidate_a9572b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e3f35f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_197a40\": 1.0, \"candidate_a9572b\": 0.0, \"candidate_e3f35f\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e310f071b0ac26401a4c9e2c:choice", "group_id": "fresh12:snli:e310f071b0ac26401a4c9e2c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man with a duffel bag and girl pushing a cart walk down a street.\\nHypothesis: A man with a duffel bag walks with agirl pushing a cart down the street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8fa7fb\": \"The hypothesis follows from the premise.\", \"candidate_34d551\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_aa4414\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8fa7fb\": 1.0, \"candidate_34d551\": 0.0, \"candidate_aa4414\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f097d639ef98fe6d0d8ef237:choice", "group_id": "fresh12:snli:f097d639ef98fe6d0d8ef237", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy and a woman are on top of large rocks at sunset.\\nHypothesis: The surface of the earth is visible.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_02a8e1\": \"The hypothesis follows from the premise.\", \"candidate_1e1e8f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9da62e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_02a8e1\": 1.0, \"candidate_1e1e8f\": 0.0, \"candidate_9da62e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9c46b2cfbcf17f8e9b419a30:choice", "group_id": "fresh12:snli:9c46b2cfbcf17f8e9b419a30", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Workers are removing ice from a walkway.\\nHypothesis: Workers outside on a slippery walkway\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_76ffb4\": \"The hypothesis follows from the premise.\", \"candidate_3b3202\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_10c53a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_76ffb4\": 1.0, \"candidate_3b3202\": 0.0, \"candidate_10c53a\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:51551154032bb513c1d180f9:choice", "group_id": "fresh12:snli:51551154032bb513c1d180f9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in a purple jacket walks by two elderly men outside a cafe.\\nHypothesis: A women walking outdoors.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_833f14\": \"The hypothesis follows from the premise.\", \"candidate_2aeeae\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c0b034\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_833f14\": 1.0, \"candidate_2aeeae\": 0.0, \"candidate_c0b034\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5ddaeef2ce7b58a1edd66ccf:choice", "group_id": "fresh12:snli:5ddaeef2ce7b58a1edd66ccf", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a grassy field throws a stick for a group of three brown dogs.\\nHypothesis: A man is playing with dogs.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8589ab\": \"The hypothesis follows from the premise.\", \"candidate_01d7e2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0e08b5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8589ab\": 1.0, \"candidate_01d7e2\": 0.0, \"candidate_0e08b5\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1760d610410d9a93a1e038a2:choice", "group_id": "fresh12:snli:1760d610410d9a93a1e038a2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man is working in an art studio and is surrounded by art supplies.\\nHypothesis: A man is working on art.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_56cd3c\": \"The hypothesis follows from the premise.\", \"candidate_053d23\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7380d0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_56cd3c\": 1.0, \"candidate_053d23\": 0.0, \"candidate_7380d0\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:87b8022de6903538495b5df0:choice", "group_id": "fresh12:snli:87b8022de6903538495b5df0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A biker wearing glass and a backpack rides near a fountain.\\nHypothesis: A biker is riding for a charity event through the center of town.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_85a887\": \"The hypothesis follows from the premise.\", \"candidate_71d350\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9f653b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_85a887\": 0.0, \"candidate_71d350\": 1.0, \"candidate_9f653b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c0c8f97288f7db4fa71d438a:choice", "group_id": "fresh12:snli:c0c8f97288f7db4fa71d438a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: At least eight people are sitting at a table in a building that overlooks the city.\\nHypothesis: people sitting at a table getting ready for dinner\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e35bbe\": \"The hypothesis follows from the premise.\", \"candidate_e8571c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_6d356c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e35bbe\": 0.0, \"candidate_e8571c\": 1.0, \"candidate_6d356c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:717eaa5050a3050df14f2945:choice", "group_id": "fresh12:snli:717eaa5050a3050df14f2945", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A dark-haired man in a black uniform is speaking to a man wearing a black jacket and black helmet on a motorbike.\\nHypothesis: A man in a police uniform is talking to another man.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f7f9e1\": \"The hypothesis follows from the premise.\", \"candidate_4a66c2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d0a297\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f7f9e1\": 0.0, \"candidate_4a66c2\": 1.0, \"candidate_d0a297\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6361aaae9cdbaf15b1faf0fb:choice", "group_id": "fresh12:snli:6361aaae9cdbaf15b1faf0fb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in an army uniform meeting someone foreign.\\nHypothesis: A man in an army uniform is meeting a foreign civilian in the war zone.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f171e2\": \"The hypothesis follows from the premise.\", \"candidate_9bb84b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_21776d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f171e2\": 0.0, \"candidate_9bb84b\": 1.0, \"candidate_21776d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4ec8c2ee47ae074e011507c1:choice", "group_id": "fresh12:snli:4ec8c2ee47ae074e011507c1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young girl in a helmet walks barefoot across a mossy tree trunk that is lying over water.\\nHypothesis: A girl might fall off a tree.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_568872\": \"The hypothesis follows from the premise.\", \"candidate_ccec08\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_dda8dd\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_568872\": 0.0, \"candidate_ccec08\": 1.0, \"candidate_dda8dd\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4a52a3c0072b91822b9e17e8:choice", "group_id": "fresh12:snli:4a52a3c0072b91822b9e17e8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Child plays with pigeon in busy public area while group of people watch him play.\\nHypothesis: A child is chasing pigeons around.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_edbc4e\": \"The hypothesis follows from the premise.\", \"candidate_e8cbcf\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ae6fbc\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_edbc4e\": 0.0, \"candidate_e8cbcf\": 1.0, \"candidate_ae6fbc\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0895fd732f5b9ad4203840df:choice", "group_id": "fresh12:snli:0895fd732f5b9ad4203840df", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two formally dressed, bald older women\\nHypothesis: Two bald women with glasses.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2c3240\": \"The hypothesis follows from the premise.\", \"candidate_6be201\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3169e3\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2c3240\": 0.0, \"candidate_6be201\": 1.0, \"candidate_3169e3\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9bc128a31cb15a66f3146ff4:choice", "group_id": "fresh12:snli:9bc128a31cb15a66f3146ff4", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man and a woman are on a seesaw.\\nHypothesis: Two attractive adults on a seesaw.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4560d8\": \"The hypothesis follows from the premise.\", \"candidate_30a25b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e25eb9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4560d8\": 0.0, \"candidate_30a25b\": 1.0, \"candidate_e25eb9\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:69c3b9b1dda78515d8fe698f:choice", "group_id": "fresh12:snli:69c3b9b1dda78515d8fe698f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: there's a woman in a pink and gray striped outfit standing with five children, three of which are making faces.\\nHypothesis: the family is trying to take a family picture\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b85b82\": \"The hypothesis follows from the premise.\", \"candidate_c3026e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ee5dce\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b85b82\": 0.0, \"candidate_c3026e\": 1.0, \"candidate_ee5dce\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:90838a78b3f3614b49bcdcdc:choice", "group_id": "fresh12:snli:90838a78b3f3614b49bcdcdc", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A brown and white dog walking away from a black and brown dog with a tennis ball in its mouth while standing in water.\\nHypothesis: Two dogs play together at the lake.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_afd5c6\": \"The hypothesis follows from the premise.\", \"candidate_f53ebe\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b33833\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_afd5c6\": 0.0, \"candidate_f53ebe\": 1.0, \"candidate_b33833\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5d469d558c8ed1f4a99e80e2:choice", "group_id": "fresh12:snli:5d469d558c8ed1f4a99e80e2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man is walking past a large sign that says E.S.E. Electronics.\\nHypothesis: The man works at E.S.E Electronics.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c36b6d\": \"The hypothesis follows from the premise.\", \"candidate_96c0bd\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_99ad09\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c36b6d\": 0.0, \"candidate_96c0bd\": 1.0, \"candidate_99ad09\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d7d1d9f2ffe1132cffa9c325:choice", "group_id": "fresh12:snli:d7d1d9f2ffe1132cffa9c325", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a navy blue shirt is tossing a boy with a large smile on his face high into the air.\\nHypothesis: A man tossing his son into the air.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4d0e0f\": \"The hypothesis follows from the premise.\", \"candidate_33319f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_51f59a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4d0e0f\": 0.0, \"candidate_33319f\": 1.0, \"candidate_51f59a\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4c0de6888b57b518df88aa81:choice", "group_id": "fresh12:snli:4c0de6888b57b518df88aa81", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Street vendors preparing and serving grilled food.\\nHypothesis: Street vendors are skillfully grilling food for customers.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0419be\": \"The hypothesis follows from the premise.\", \"candidate_9a425b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3145b1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0419be\": 0.0, \"candidate_9a425b\": 1.0, \"candidate_3145b1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:8e9bfe83c146182617a7b8c2:choice", "group_id": "fresh12:snli:8e9bfe83c146182617a7b8c2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Four young boys flexing for the camera.\\nHypothesis: They are friends\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_765d37\": \"The hypothesis follows from the premise.\", \"candidate_135280\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ea7de7\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_765d37\": 0.0, \"candidate_135280\": 1.0, \"candidate_ea7de7\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:25a0e058a4290aed2918b180:choice", "group_id": "fresh12:snli:25a0e058a4290aed2918b180", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man wearing black carving stone.\\nHypothesis: The man is very tall.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a97d22\": \"The hypothesis follows from the premise.\", \"candidate_f07652\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_fe2271\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a97d22\": 0.0, \"candidate_f07652\": 1.0, \"candidate_fe2271\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d225ce2a2bd286f6648b5946:choice", "group_id": "fresh12:snli:d225ce2a2bd286f6648b5946", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman outside in the snow holds a container that produces steam.\\nHypothesis: A women about to make hot water turn into snow by throwing the hot water in the air.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8a51af\": \"The hypothesis follows from the premise.\", \"candidate_c66fb7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a98e1f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8a51af\": 0.0, \"candidate_c66fb7\": 1.0, \"candidate_a98e1f\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5a0462a620fe157ebda36911:choice", "group_id": "fresh12:snli:5a0462a620fe157ebda36911", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A little boy in a red shirt and swim trunks plays at the beach.\\nHypothesis: The boy is having fun.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7cd116\": \"The hypothesis follows from the premise.\", \"candidate_a52fce\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5c1d48\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7cd116\": 0.0, \"candidate_a52fce\": 1.0, \"candidate_5c1d48\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c0e8e02e9aadfe8530ac590a:choice", "group_id": "fresh12:snli:c0e8e02e9aadfe8530ac590a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Navy men stand along a red carpet and salute as their superior walks by.\\nHypothesis: Some men are saluting their superior out of respect.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5ccbb4\": \"The hypothesis follows from the premise.\", \"candidate_76277e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d012f5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5ccbb4\": 0.0, \"candidate_76277e\": 1.0, \"candidate_d012f5\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6a64c880c3a74015a6223b8f:choice", "group_id": "fresh12:snli:6a64c880c3a74015a6223b8f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man with medium length curly hair, wearing a button down shirt and bracelets, sings into a microphone while he plays a guitar.\\nHypothesis: A tall person with hair\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b0ea2c\": \"The hypothesis follows from the premise.\", \"candidate_96570d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_70ffa4\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b0ea2c\": 0.0, \"candidate_96570d\": 1.0, \"candidate_70ffa4\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ce6107ef3ca1891ba72c6ecf:choice", "group_id": "fresh12:snli:ce6107ef3ca1891ba72c6ecf", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A large number of people wearing only white and red are walking on a street.\\nHypothesis: A large number of men walking down the street in red and white.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dcafb9\": \"The hypothesis follows from the premise.\", \"candidate_6592ae\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8f3f06\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dcafb9\": 0.0, \"candidate_6592ae\": 1.0, \"candidate_8f3f06\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:95774babeeb0609e73bf59bc:choice", "group_id": "fresh12:snli:95774babeeb0609e73bf59bc", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: One young woman in a blue uniform helping another young woman in a blue sweatshirt with a white band around her arm.\\nHypothesis: A tall person in a uniform\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_367398\": \"The hypothesis follows from the premise.\", \"candidate_0696dc\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c64e7e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_367398\": 0.0, \"candidate_0696dc\": 1.0, \"candidate_c64e7e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6d0dc1343f7c6170be3ff2e6:choice", "group_id": "fresh12:snli:6d0dc1343f7c6170be3ff2e6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a lady looks surprised.\\nHypothesis: A woman looks surprised because someone just dropped to one knee and proposed!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_99542f\": \"The hypothesis follows from the premise.\", \"candidate_3c7826\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_090d63\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_99542f\": 0.0, \"candidate_3c7826\": 1.0, \"candidate_090d63\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:3e55bb27b7605358d32aaecb:choice", "group_id": "fresh12:snli:3e55bb27b7605358d32aaecb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Man taking photos of seagulls over the ocean.\\nHypothesis: It is sunny.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8e00cd\": \"The hypothesis follows from the premise.\", \"candidate_4900e2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3b7f17\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8e00cd\": 0.0, \"candidate_4900e2\": 1.0, \"candidate_3b7f17\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5bf7a6965ca95d6574c7a8de:choice", "group_id": "fresh12:snli:5bf7a6965ca95d6574c7a8de", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Woman in a blue dress and carrying a black purse, rides a scooter down the street.\\nHypothesis: the girl is riding down a dirt street\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_735813\": \"The hypothesis follows from the premise.\", \"candidate_9a82a5\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_6a8be0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_735813\": 0.0, \"candidate_9a82a5\": 1.0, \"candidate_6a8be0\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d2bdd826588e34d5ad812958:choice", "group_id": "fresh12:snli:d2bdd826588e34d5ad812958", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man with a tattoo on his arm staring to the side with vehicles and buildings behind him.\\nHypothesis: A man has dragon tattoos on his arm.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3b9f46\": \"The hypothesis follows from the premise.\", \"candidate_e0cdde\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ccbef9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3b9f46\": 0.0, \"candidate_e0cdde\": 1.0, \"candidate_ccbef9\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b835ef345ac41abb0670d919:choice", "group_id": "fresh12:snli:b835ef345ac41abb0670d919", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two small boys dressed in tuxedos sitting on a red carpeted floor.\\nHypothesis: Two well dressed boys sit on a red carpet floor inside of a home.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_65489b\": \"The hypothesis follows from the premise.\", \"candidate_81781f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b93bc8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_65489b\": 0.0, \"candidate_81781f\": 1.0, \"candidate_b93bc8\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ef68a1d6e8a76e665e7bcfbd:choice", "group_id": "fresh12:snli:ef68a1d6e8a76e665e7bcfbd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A couple about to kiss somewhere.\\nHypothesis: A couple are kissing on new years eve as the ball drops.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ec3417\": \"The hypothesis follows from the premise.\", \"candidate_cdb921\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_841868\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ec3417\": 0.0, \"candidate_cdb921\": 1.0, \"candidate_841868\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5e5b40e35de48c1433ea4a73:choice", "group_id": "fresh12:snli:5e5b40e35de48c1433ea4a73", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A yellow race car sliding through a corner as spectators watch.\\nHypothesis: A Nascar is being watched.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c1466d\": \"The hypothesis follows from the premise.\", \"candidate_02d00c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4b9608\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c1466d\": 0.0, \"candidate_02d00c\": 1.0, \"candidate_4b9608\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e47a95c4b57d7372e7648989:choice", "group_id": "fresh12:snli:e47a95c4b57d7372e7648989", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boat pear with people boarding and disembarking some boats.\\nHypothesis: people are starting and leaving a fishing boat tour\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6cde63\": \"The hypothesis follows from the premise.\", \"candidate_9eba58\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_6dd111\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6cde63\": 0.0, \"candidate_9eba58\": 1.0, \"candidate_6dd111\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:633bce784132a8e9cc6e9dbd:choice", "group_id": "fresh12:snli:633bce784132a8e9cc6e9dbd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A half dressed man is sleeping on the sidewalk.\\nHypothesis: A homeless man is sleeping outside\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e2c33c\": \"The hypothesis follows from the premise.\", \"candidate_6b7bf4\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b00f49\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e2c33c\": 0.0, \"candidate_6b7bf4\": 1.0, \"candidate_b00f49\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f2581330cbd6701c0734064a:choice", "group_id": "fresh12:snli:f2581330cbd6701c0734064a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: People waiting to cross the street on the side of a sidewalk.\\nHypothesis: People are crossing the street during a parade.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7970d4\": \"The hypothesis follows from the premise.\", \"candidate_4be4e1\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_fcf6b3\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7970d4\": 0.0, \"candidate_4be4e1\": 1.0, \"candidate_fcf6b3\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2e0d159326aaeeda9ecf84af:choice", "group_id": "fresh12:snli:2e0d159326aaeeda9ecf84af", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A female gymnast flies off of the lower bar and is suspended in the air with her feet pointed upward.\\nHypothesis: A female gymnast is competing at the Olympics.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1a796d\": \"The hypothesis follows from the premise.\", \"candidate_26d0a3\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_75bdbf\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1a796d\": 0.0, \"candidate_26d0a3\": 1.0, \"candidate_75bdbf\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:342e1f10a6ebafc6b09c9b8d:choice", "group_id": "fresh12:snli:342e1f10a6ebafc6b09c9b8d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An African American man is posing in front of an E.S.E. Electronics advertisement billboard after purchasing items from the local liquor store.\\nHypothesis: Man takes a picture with advertiser after purchasing their brand of liquor at a store they are promoting.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bfb5d4\": \"The hypothesis follows from the premise.\", \"candidate_41b93b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_772f0e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bfb5d4\": 0.0, \"candidate_41b93b\": 1.0, \"candidate_772f0e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7c44aba080f5912d9dc07053:choice", "group_id": "fresh12:snli:7c44aba080f5912d9dc07053", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Six males, four with shirts and two without, have taken a break from their construction work on a building.\\nHypothesis: A construction crew is sitting and standing next to a building.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4fa2b4\": \"The hypothesis follows from the premise.\", \"candidate_ae5a02\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ccd497\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4fa2b4\": 0.0, \"candidate_ae5a02\": 1.0, \"candidate_ccd497\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4ed21da2fb143ce3324c16e1:choice", "group_id": "fresh12:snli:4ed21da2fb143ce3324c16e1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two women bend over to touch a small kangaroo.\\nHypothesis: The kangaroo wants to leave\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_59e3db\": \"The hypothesis follows from the premise.\", \"candidate_69c777\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_6a0e91\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_59e3db\": 0.0, \"candidate_69c777\": 1.0, \"candidate_6a0e91\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b88ac26189fe94f134d15518:choice", "group_id": "fresh12:snli:b88ac26189fe94f134d15518", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young man appears bored in a restaurant.\\nHypothesis: A young sad man appears bored in a restaurant.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ed3f71\": \"The hypothesis follows from the premise.\", \"candidate_de6688\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e1ae30\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ed3f71\": 0.0, \"candidate_de6688\": 1.0, \"candidate_e1ae30\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cc6e48e8c36d65d401a24c89:choice", "group_id": "fresh12:snli:cc6e48e8c36d65d401a24c89", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Kids play in water coming up in streams out of a tiled floor with an image of a large rose on it.\\nHypothesis: Kids are playing at the fountain at a local mall.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_426160\": \"The hypothesis follows from the premise.\", \"candidate_a07e08\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f6d216\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_426160\": 0.0, \"candidate_a07e08\": 1.0, \"candidate_f6d216\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0ba1c5eb9e120b2d8206daba:choice", "group_id": "fresh12:snli:0ba1c5eb9e120b2d8206daba", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An elderly man is smiling while sitting in front of a row of soda cans.\\nHypothesis: An old man collects soda cans.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f358ac\": \"The hypothesis follows from the premise.\", \"candidate_1911e8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_fc7910\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f358ac\": 0.0, \"candidate_1911e8\": 1.0, \"candidate_fc7910\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a175fd0e7dd5e65305e2545e:choice", "group_id": "fresh12:snli:a175fd0e7dd5e65305e2545e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men on stilts and a woman twirling a ribbon, lead a festive parade along the streets, while a woman looks on from her window.\\nHypothesis: Two men are on stilts while a woman dances between them in the parade.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4c7107\": \"The hypothesis follows from the premise.\", \"candidate_a7ed1d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_850a38\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4c7107\": 0.0, \"candidate_a7ed1d\": 1.0, \"candidate_850a38\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2efa09a3036e3ad3259f49f8:choice", "group_id": "fresh12:snli:2efa09a3036e3ad3259f49f8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: There are two kilted men, one of them older and is holding bagpipes with the other one with a drum.\\nHypothesis: Two men are friends\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_623e19\": \"The hypothesis follows from the premise.\", \"candidate_93f4ee\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_cbc156\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_623e19\": 0.0, \"candidate_93f4ee\": 1.0, \"candidate_cbc156\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a16dd71857f9b653db9dae67:choice", "group_id": "fresh12:snli:a16dd71857f9b653db9dae67", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man on a bicycle, wearing cycle gear, riding at a fast past down paved trail surrounded by tree's and grass.\\nHypothesis: He is heading to the cabin down the trail.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_607e96\": \"The hypothesis follows from the premise.\", \"candidate_19bd39\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b75b40\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_607e96\": 0.0, \"candidate_19bd39\": 1.0, \"candidate_b75b40\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7756b2ce440f0db8d420ffad:choice", "group_id": "fresh12:snli:7756b2ce440f0db8d420ffad", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in an office making a phone call\\nHypothesis: A woman is making a business phone call.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_de66d7\": \"The hypothesis follows from the premise.\", \"candidate_03125a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5bdfcc\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_de66d7\": 0.0, \"candidate_03125a\": 1.0, \"candidate_5bdfcc\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:44d51d98b55e2286c894c43c:choice", "group_id": "fresh12:snli:44d51d98b55e2286c894c43c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man and a woman walk at a flea market.\\nHypothesis: The couple are shopping for deals.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2f0000\": \"The hypothesis follows from the premise.\", \"candidate_3fe823\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_57aace\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2f0000\": 0.0, \"candidate_3fe823\": 1.0, \"candidate_57aace\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5b29dbca6e8573449dd1703a:choice", "group_id": "fresh12:snli:5b29dbca6e8573449dd1703a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Some elderly people are playing instruments in a park.\\nHypothesis: The old man is playing the guitar.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_96d707\": \"The hypothesis follows from the premise.\", \"candidate_518485\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7b5c7d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_96d707\": 0.0, \"candidate_518485\": 1.0, \"candidate_7b5c7d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ec1faf5899f700a8aeb7ac42:choice", "group_id": "fresh12:snli:ec1faf5899f700a8aeb7ac42", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A red truck is parked next to a burning blue building while a man in a green vest runs toward it.\\nHypothesis: The burning blue building will turn to dust after an hour.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a3867c\": \"The hypothesis follows from the premise.\", \"candidate_fe421d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_528f11\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a3867c\": 0.0, \"candidate_fe421d\": 1.0, \"candidate_528f11\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a10e97e43b833000dcd493b2:choice", "group_id": "fresh12:snli:a10e97e43b833000dcd493b2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: While traveling, a woman in a black jacket is resting on the shoulder of a man in a black jacket.\\nHypothesis: A man's wife lays on his shoulder\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1bbbf4\": \"The hypothesis follows from the premise.\", \"candidate_79a6c2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_52031d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1bbbf4\": 0.0, \"candidate_79a6c2\": 1.0, \"candidate_52031d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6697c2a30584a1d448fc5b9f:choice", "group_id": "fresh12:snli:6697c2a30584a1d448fc5b9f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Man with blue and yellow shirt rides bike.\\nHypothesis: a bike rides a man for its first time\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0d79e3\": \"The hypothesis follows from the premise.\", \"candidate_a593ea\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_040ae9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0d79e3\": 0.0, \"candidate_a593ea\": 1.0, \"candidate_040ae9\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f304be836969ff15919566be:choice", "group_id": "fresh12:snli:f304be836969ff15919566be", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Military personnel are shopping\\nHypothesis: Military personnel are buying groceries.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6ebdee\": \"The hypothesis follows from the premise.\", \"candidate_332b88\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f58694\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6ebdee\": 0.0, \"candidate_332b88\": 1.0, \"candidate_f58694\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:086c99ecc0ebbcb1c1266c3d:choice", "group_id": "fresh12:snli:086c99ecc0ebbcb1c1266c3d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three Asian men are running down a track holding hands while a large gathering of people watches them.\\nHypothesis: Three men are performing a ritual.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9172a6\": \"The hypothesis follows from the premise.\", \"candidate_caf5dd\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b5384f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9172a6\": 0.0, \"candidate_caf5dd\": 1.0, \"candidate_b5384f\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4683ddf2b3d8b20c86636953:choice", "group_id": "fresh12:snli:4683ddf2b3d8b20c86636953", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two older men in business attire are walking in a city, text graffiti is visible on the sidewalk.\\nHypothesis: They hope to clean up the street together\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1bc285\": \"The hypothesis follows from the premise.\", \"candidate_a9b695\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e65bdf\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1bc285\": 0.0, \"candidate_a9b695\": 1.0, \"candidate_e65bdf\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0b310c8f3857e5b5d160c135:choice", "group_id": "fresh12:snli:0b310c8f3857e5b5d160c135", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three soccer players, two in orange one in yellow, running for the ball on a soccer field.\\nHypothesis: Some teams were playing ball in the field.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ecb3bb\": \"The hypothesis follows from the premise.\", \"candidate_8401fa\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8130c0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ecb3bb\": 0.0, \"candidate_8401fa\": 1.0, \"candidate_8130c0\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:252342b4ef7b208598f766fd:choice", "group_id": "fresh12:snli:252342b4ef7b208598f766fd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A softball player throws the ball to her teammate.\\nHypothesis: Teammates are competing in a softballs championship.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7241ef\": \"The hypothesis follows from the premise.\", \"candidate_00e886\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9b54fb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7241ef\": 0.0, \"candidate_00e886\": 1.0, \"candidate_9b54fb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cb4125da50b79c637a3cdda9:choice", "group_id": "fresh12:snli:cb4125da50b79c637a3cdda9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men are fixing the side of an old brick building.\\nHypothesis: Construction workers at a job.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9fada0\": \"The hypothesis follows from the premise.\", \"candidate_f33489\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a30934\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9fada0\": 0.0, \"candidate_f33489\": 1.0, \"candidate_a30934\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a5774578b4244c4949330484:choice", "group_id": "fresh12:snli:a5774578b4244c4949330484", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: batter hits a baseball\\nHypothesis: A man at home plate hits a baseball with the bat.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cd2d2d\": \"The hypothesis follows from the premise.\", \"candidate_838179\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c8876a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cd2d2d\": 0.0, \"candidate_838179\": 1.0, \"candidate_c8876a\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2b395d0a324e1cddcf160c37:choice", "group_id": "fresh12:snli:2b395d0a324e1cddcf160c37", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A person wearing a blue backpack stands on a field above a scenic view.\\nHypothesis: The person is taking a nature hike.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_25c28b\": \"The hypothesis follows from the premise.\", \"candidate_58adcb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_43e00c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_25c28b\": 0.0, \"candidate_58adcb\": 1.0, \"candidate_43e00c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:fa04d5c7c58d25209423dfb9:choice", "group_id": "fresh12:snli:fa04d5c7c58d25209423dfb9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A fisherman using a cellphone on a boat.\\nHypothesis: A fisherman uses a cell phone to call his wife.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_776baa\": \"The hypothesis follows from the premise.\", \"candidate_be6efe\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_56b3bc\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_776baa\": 0.0, \"candidate_be6efe\": 1.0, \"candidate_56b3bc\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cadf1de48dadfb2434fd1e1a:choice", "group_id": "fresh12:snli:cadf1de48dadfb2434fd1e1a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy wearing jeans, standing on rocks in a river and throwing a stone.\\nHypothesis: A boy with no shirt on throws rocks.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eab93b\": \"The hypothesis follows from the premise.\", \"candidate_25be46\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d7f68e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eab93b\": 0.0, \"candidate_25be46\": 1.0, \"candidate_d7f68e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f3626ebb3b69beb7268ad1f2:choice", "group_id": "fresh12:snli:f3626ebb3b69beb7268ad1f2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A person is sitting on a yellow picnic table with a water bottle, while a baby stroller sits beside her.\\nHypothesis: a mom takes a break in a park\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d8447b\": \"The hypothesis follows from the premise.\", \"candidate_73b47a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a12fed\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d8447b\": 0.0, \"candidate_73b47a\": 1.0, \"candidate_a12fed\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6bac5575a3dc2621e75c0628:choice", "group_id": "fresh12:snli:6bac5575a3dc2621e75c0628", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: First Lady Laura Bush at podium, in front of seated audience, at the White House Conference on Global Literacy.\\nHypothesis: Laura Bush gets ready to address a heated issue\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bad9ee\": \"The hypothesis follows from the premise.\", \"candidate_c65088\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b7533a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bad9ee\": 0.0, \"candidate_c65088\": 1.0, \"candidate_b7533a\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0a6d043ae8dc573e3fb8d2c4:choice", "group_id": "fresh12:snli:0a6d043ae8dc573e3fb8d2c4", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young boy in a straw hat drinks out of a plastic pink cup.\\nHypothesis: A kid is holding a soda can and walking through the park.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e0930d\": \"The hypothesis follows from the premise.\", \"candidate_277acb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2b0079\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e0930d\": 0.0, \"candidate_277acb\": 1.0, \"candidate_2b0079\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c89412b4cc1b5dc3c7773c2c:choice", "group_id": "fresh12:snli:c89412b4cc1b5dc3c7773c2c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The young man is sitting on a couch smoking a cigarette.\\nHypothesis: A young man smoking a cigar\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e46667\": \"The hypothesis follows from the premise.\", \"candidate_9d11b0\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e0bb6c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e46667\": 0.0, \"candidate_9d11b0\": 1.0, \"candidate_e0bb6c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4c0fa17276af55936c66e67b:choice", "group_id": "fresh12:snli:4c0fa17276af55936c66e67b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Men at a party hanging out on the front porch.\\nHypothesis: They are smoking cigarettes on the porch.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_758730\": \"The hypothesis follows from the premise.\", \"candidate_a42d86\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4132cf\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_758730\": 0.0, \"candidate_a42d86\": 1.0, \"candidate_4132cf\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c6ef790cd8446a46f055a18b:choice", "group_id": "fresh12:snli:c6ef790cd8446a46f055a18b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man stands with his arms out inside a large green piece of equipment.\\nHypothesis: He is trying to get it working again.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_963008\": \"The hypothesis follows from the premise.\", \"candidate_001eab\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0d124d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_963008\": 0.0, \"candidate_001eab\": 1.0, \"candidate_0d124d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:83955c3bd7d06577f4a27774:choice", "group_id": "fresh12:snli:83955c3bd7d06577f4a27774", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Four children are sitting at a table eating breakfast.\\nHypothesis: Children are eating cereal.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7fa4fc\": \"The hypothesis follows from the premise.\", \"candidate_d2e215\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5166fa\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7fa4fc\": 0.0, \"candidate_d2e215\": 1.0, \"candidate_5166fa\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:df2171e248d668dab4939364:choice", "group_id": "fresh12:snli:df2171e248d668dab4939364", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Many people standing outside of a place talking to each other in front of a building that has a sign that says \\\"HI-POINTE.\\\"\\nHypothesis: The people are having a chat before going into the work building.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_decbef\": \"The hypothesis follows from the premise.\", \"candidate_434408\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c8f285\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_decbef\": 0.0, \"candidate_434408\": 1.0, \"candidate_c8f285\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:3b9d9adbe7c925672fb25bbe:choice", "group_id": "fresh12:snli:3b9d9adbe7c925672fb25bbe", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman wearing glasses and a brown beanie next to a girl with long brown hair holding a book.\\nHypothesis: A woman is wearing glasses and a light brown beanie.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_de5aae\": \"The hypothesis follows from the premise.\", \"candidate_5f77b8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4935cf\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_de5aae\": 0.0, \"candidate_5f77b8\": 1.0, \"candidate_4935cf\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9548c020ef732fb64e7a3324:choice", "group_id": "fresh12:snli:9548c020ef732fb64e7a3324", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Naked children play on the muddy riverbank.\\nHypothesis: THE NUDE KIDS PLAYED ON THE RIVERBANK  IN THE MUD JUST TO HAVE A MUD BATH\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_71387c\": \"The hypothesis follows from the premise.\", \"candidate_913c64\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ad1221\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_71387c\": 0.0, \"candidate_913c64\": 1.0, \"candidate_ad1221\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:aa3ed74e665fa461ab197ed5:choice", "group_id": "fresh12:snli:aa3ed74e665fa461ab197ed5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two young women, one with blond-hair in a ponytail, are sitting in a booth next to some potted plants.\\nHypothesis: Two women are sitting near to green plants\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_283493\": \"The hypothesis follows from the premise.\", \"candidate_b47334\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_157972\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_283493\": 0.0, \"candidate_b47334\": 1.0, \"candidate_157972\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:72ade2555b493d0a908b77d7:choice", "group_id": "fresh12:snli:72ade2555b493d0a908b77d7", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A shoeshine man works as others pass by on a busy sidewalk.\\nHypothesis: A man is almost done working today.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_430418\": \"The hypothesis follows from the premise.\", \"candidate_27c9bc\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f26b88\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_430418\": 0.0, \"candidate_27c9bc\": 1.0, \"candidate_f26b88\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bb3347f5e561d3bc201d424e:choice", "group_id": "fresh12:snli:bb3347f5e561d3bc201d424e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The small child is wearing winter clothes and a hat.\\nHypothesis: The small child is dressed in a parka.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e7ccf4\": \"The hypothesis follows from the premise.\", \"candidate_3fdeef\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_6d54ee\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e7ccf4\": 0.0, \"candidate_3fdeef\": 1.0, \"candidate_6d54ee\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5b1b73ccdbc87a5727c85543:choice", "group_id": "fresh12:snli:5b1b73ccdbc87a5727c85543", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Many people who appear to be Asian stand in line at a train station waiting patiently for the next train.\\nHypothesis: some people standing\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_066c11\": \"The hypothesis follows from the premise.\", \"candidate_1d90e0\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_42a6bd\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_066c11\": 0.0, \"candidate_1d90e0\": 1.0, \"candidate_42a6bd\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0ea3efaf8d07cca9bebf1bba:choice", "group_id": "fresh12:snli:0ea3efaf8d07cca9bebf1bba", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The German shepherd runs happily in the park.\\nHypothesis: Dog runs outside with his owner.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_75b055\": \"The hypothesis follows from the premise.\", \"candidate_6855da\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7d9915\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_75b055\": 0.0, \"candidate_6855da\": 1.0, \"candidate_7d9915\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:85eddb1e2a39deb62f105767:choice", "group_id": "fresh12:snli:85eddb1e2a39deb62f105767", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men have an after dinner discussion at the dinner table.\\nHypothesis: the men are partners\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3152e4\": \"The hypothesis follows from the premise.\", \"candidate_458e3a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3af52f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3152e4\": 0.0, \"candidate_458e3a\": 1.0, \"candidate_3af52f\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b8eedd162cf778bf16e82500:choice", "group_id": "fresh12:snli:b8eedd162cf778bf16e82500", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: This person in a yellow kayak is riding the rapids.\\nHypothesis: This person is going to the waterfall\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7d4f97\": \"The hypothesis follows from the premise.\", \"candidate_00b395\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5c7846\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7d4f97\": 0.0, \"candidate_00b395\": 1.0, \"candidate_5c7846\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a2b6eafe146644c6ee516a19:choice", "group_id": "fresh12:snli:a2b6eafe146644c6ee516a19", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman is sitting in a chair, while a man is going through the sport utility vehicle.\\nHypothesis: a girl is sitting waiting to leave for vacation\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a477d2\": \"The hypothesis follows from the premise.\", \"candidate_49e953\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_12a60d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a477d2\": 0.0, \"candidate_49e953\": 1.0, \"candidate_12a60d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b711add33aca49c2f0f3252a:choice", "group_id": "fresh12:snli:b711add33aca49c2f0f3252a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men are dressed up as snowmen.\\nHypothesis: two young men are dressed up\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7b9635\": \"The hypothesis follows from the premise.\", \"candidate_5a432b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_78aaa1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7b9635\": 0.0, \"candidate_5a432b\": 1.0, \"candidate_78aaa1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:952ae35e4635183b22f63ccf:choice", "group_id": "fresh12:snli:952ae35e4635183b22f63ccf", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a plaid shirt looking through a telescope lens.\\nHypothesis: the man is watching the stars\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_af26ed\": \"The hypothesis follows from the premise.\", \"candidate_7c7a51\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c967d3\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_af26ed\": 0.0, \"candidate_7c7a51\": 1.0, \"candidate_c967d3\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:df2433d372e81ace5a93530b:choice", "group_id": "fresh12:snli:df2433d372e81ace5a93530b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A casually dressed young blond woman is looking away sadly as couple near her are kissing.\\nHypothesis: A woman looks away after recognizing her ex-boyfriend kissing his new girl.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ae3cba\": \"The hypothesis follows from the premise.\", \"candidate_34f65f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1b7f48\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ae3cba\": 0.0, \"candidate_34f65f\": 1.0, \"candidate_1b7f48\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1c8c20b08434eb1e3803a713:choice", "group_id": "fresh12:snli:1c8c20b08434eb1e3803a713", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young man in a gray shirt hovers over his beer bottle and Guinness glass pulling green things out of his beard.\\nHypothesis: The man in the gray t-shirt pulls little green men out of beard; too much drinking, I guess.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_718f78\": \"The hypothesis follows from the premise.\", \"candidate_403c85\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_79f40a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_718f78\": 0.0, \"candidate_403c85\": 1.0, \"candidate_79f40a\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:92f1dfdf6676a406082a1c56:choice", "group_id": "fresh12:snli:92f1dfdf6676a406082a1c56", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man waiting for the bus in a war torn country.\\nHypothesis: A man is waiting for a bus sitting on a bench.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_400fc9\": \"The hypothesis follows from the premise.\", \"candidate_efc228\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_24bd5d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_400fc9\": 0.0, \"candidate_efc228\": 1.0, \"candidate_24bd5d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:13ef2bffd7165eceeb0adaf3:choice", "group_id": "fresh12:snli:13ef2bffd7165eceeb0adaf3", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Motorcycle rider going over a bump.\\nHypothesis: Motorcycle rider going over a big bump.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_24caca\": \"The hypothesis follows from the premise.\", \"candidate_525419\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_aed909\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_24caca\": 0.0, \"candidate_525419\": 1.0, \"candidate_aed909\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:700419f04b2acef40f608c52:choice", "group_id": "fresh12:snli:700419f04b2acef40f608c52", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: There is a person walking down a tree lined street with snow scattered about.\\nHypothesis: There is a man walking down a tree lined street with snow scattered about.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_baf077\": \"The hypothesis follows from the premise.\", \"candidate_c98479\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_62531c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_baf077\": 0.0, \"candidate_c98479\": 1.0, \"candidate_62531c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:af407434231d5beb8756876c:choice", "group_id": "fresh12:snli:af407434231d5beb8756876c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Eight tourists stand at a railing looking at an attraction.\\nHypothesis: The tourists were on a bus tour\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_080bbe\": \"The hypothesis follows from the premise.\", \"candidate_f06586\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7c9ed0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_080bbe\": 0.0, \"candidate_f06586\": 1.0, \"candidate_7c9ed0\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b3e9a3a07c62a0bed3de08c0:choice", "group_id": "fresh12:snli:b3e9a3a07c62a0bed3de08c0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man and woman talk to another woman, all looking concerned, in a room being decorated with streamers and balloons.\\nHypothesis: Some people are throwing a surprise birthday party.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0e3804\": \"The hypothesis follows from the premise.\", \"candidate_6de0c6\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_091b82\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0e3804\": 0.0, \"candidate_6de0c6\": 1.0, \"candidate_091b82\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:41cc012e75b764b7a5b33b01:choice", "group_id": "fresh12:snli:41cc012e75b764b7a5b33b01", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three people are making transactions at an automated teller machine.\\nHypothesis: Three women waiting for an ATM.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3b61cb\": \"The hypothesis follows from the premise.\", \"candidate_8c286a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_81a871\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3b61cb\": 0.0, \"candidate_8c286a\": 1.0, \"candidate_81a871\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:84ec52d262747984f7640fa1:choice", "group_id": "fresh12:snli:84ec52d262747984f7640fa1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Am man riding a very tall bicycle down the street, past a large brick building.\\nHypothesis: A clown rides a tall bike in a parade.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e0cc9a\": \"The hypothesis follows from the premise.\", \"candidate_195996\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0ae9c9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e0cc9a\": 0.0, \"candidate_195996\": 1.0, \"candidate_0ae9c9\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7ec27ecbd0d831f036794d98:choice", "group_id": "fresh12:snli:7ec27ecbd0d831f036794d98", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man is standing inside a building under construction filled with bricks and various tools.\\nHypothesis: a contractor looks at the job\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_974267\": \"The hypothesis follows from the premise.\", \"candidate_4c370a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b9ae63\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_974267\": 0.0, \"candidate_4c370a\": 1.0, \"candidate_b9ae63\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:117409ee3d820bf7945017f3:choice", "group_id": "fresh12:snli:117409ee3d820bf7945017f3", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two people cooking a meal together in a household kitchen.\\nHypothesis: The people are cooking a dinner for Valentine's Day.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d13c13\": \"The hypothesis follows from the premise.\", \"candidate_4fe91e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1ea643\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d13c13\": 0.0, \"candidate_4fe91e\": 1.0, \"candidate_1ea643\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2ecdbe8a0ef3b85d64a73de5:choice", "group_id": "fresh12:snli:2ecdbe8a0ef3b85d64a73de5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A little girl has her arms around a little boy standing on a wooden bridge in the woods.\\nHypothesis: The little girl is consoling the boy.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8d660c\": \"The hypothesis follows from the premise.\", \"candidate_91ba94\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c9bb2c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8d660c\": 0.0, \"candidate_91ba94\": 1.0, \"candidate_c9bb2c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:336eba5e3cc530aeb7798e51:choice", "group_id": "fresh12:snli:336eba5e3cc530aeb7798e51", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men are in blue life jackets and sitting on innertubes with \\\"Crusher\\\" written on them in the middle of a body of water.\\nHypothesis: The two men are enjoying an afternoon in the water together after a tough workweek.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d6f57f\": \"The hypothesis follows from the premise.\", \"candidate_20a453\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5e68e3\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d6f57f\": 0.0, \"candidate_20a453\": 1.0, \"candidate_5e68e3\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0a28ae970a5f4a7024842234:choice", "group_id": "fresh12:snli:0a28ae970a5f4a7024842234", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A shirtless man with red paint on his torso holding a stick.\\nHypothesis: A rabid sports fan cheers for his team.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_85f987\": \"The hypothesis follows from the premise.\", \"candidate_2d4482\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9f635d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_85f987\": 0.0, \"candidate_2d4482\": 1.0, \"candidate_9f635d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:3ad201c493d729abf7513d07:choice", "group_id": "fresh12:snli:3ad201c493d729abf7513d07", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Several members of a community are celebrating with music and costumes during a festival or party.\\nHypothesis: The people are celebrating New Years Eve.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_991fcc\": \"The hypothesis follows from the premise.\", \"candidate_aa0fd1\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_94c202\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_991fcc\": 0.0, \"candidate_aa0fd1\": 1.0, \"candidate_94c202\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0bfaeb5c603488b0b0292f76:choice", "group_id": "fresh12:snli:0bfaeb5c603488b0b0292f76", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Children are playing outside in a fountain.\\nHypothesis: Some children a outside playing near the water.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f974fd\": \"The hypothesis follows from the premise.\", \"candidate_c9349c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_41c5be\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f974fd\": 0.0, \"candidate_c9349c\": 1.0, \"candidate_41c5be\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:517c42e2c5a6758283950413:choice", "group_id": "fresh12:snli:517c42e2c5a6758283950413", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A solo shot of a man riding a four wheeler in an atv event.\\nHypothesis: The man is riding fast on the four wheeler.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a7f92e\": \"The hypothesis follows from the premise.\", \"candidate_e3321b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_83f3f8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a7f92e\": 0.0, \"candidate_e3321b\": 1.0, \"candidate_83f3f8\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5548f3d615ceebe3fac15a60:choice", "group_id": "fresh12:snli:5548f3d615ceebe3fac15a60", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Cityscape with four men on bikes serving as the focal point of picture.\\nHypothesis: Bike rental as low as 9.99 an hour.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c02e44\": \"The hypothesis follows from the premise.\", \"candidate_d9ff92\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2ae717\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c02e44\": 0.0, \"candidate_d9ff92\": 1.0, \"candidate_2ae717\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ab7a94c4a2813a755150d420:choice", "group_id": "fresh12:snli:ab7a94c4a2813a755150d420", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman is holding a large check for Kids Food Basket.\\nHypothesis: She is making a donation.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_017b5b\": \"The hypothesis follows from the premise.\", \"candidate_c2bb94\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_26cb20\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_017b5b\": 0.0, \"candidate_c2bb94\": 1.0, \"candidate_26cb20\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:56dda6641efe760f7a89ddc6:choice", "group_id": "fresh12:snli:56dda6641efe760f7a89ddc6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman standing in front of a wall holding a can of Bud Light with her shirt tied into a semi-halter.\\nHypothesis: The woman who is shown lives in the South.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8d89ac\": \"The hypothesis follows from the premise.\", \"candidate_debe42\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2d2281\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8d89ac\": 0.0, \"candidate_debe42\": 1.0, \"candidate_2d2281\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:fd999926c668e0fc749b38c9:choice", "group_id": "fresh12:snli:fd999926c668e0fc749b38c9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman within an orchestra is playing a violin.\\nHypothesis: A woman is playing a concert.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_42d0ce\": \"The hypothesis follows from the premise.\", \"candidate_49530e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c057b7\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_42d0ce\": 0.0, \"candidate_49530e\": 1.0, \"candidate_c057b7\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:63ac5a5825ea6135eb321b75:choice", "group_id": "fresh12:snli:63ac5a5825ea6135eb321b75", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy wearing a shirt with the Simpsons character, Ralph Wiggum, plays a Nintendo DS with his sister.\\nHypothesis: A boy is playing a Nintendo DS with his sister who he loves.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_442739\": \"The hypothesis follows from the premise.\", \"candidate_751b53\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_592440\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_442739\": 0.0, \"candidate_751b53\": 1.0, \"candidate_592440\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5ae3ce107262e1b80690f839:choice", "group_id": "fresh12:snli:5ae3ce107262e1b80690f839", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A road crew is working on a road with equipment around them.\\nHypothesis: A road crew in orange vests are working on a road with equipment around them.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c022be\": \"The hypothesis follows from the premise.\", \"candidate_e8ebaa\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_76480e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c022be\": 0.0, \"candidate_e8ebaa\": 1.0, \"candidate_76480e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c5a28187795e9bd8a96fc8ba:choice", "group_id": "fresh12:snli:c5a28187795e9bd8a96fc8ba", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The pitcher is pitching the ball in a game of baseball.\\nHypothesis: The pitcher wants to strike out the last player on the team.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_819b1a\": \"The hypothesis follows from the premise.\", \"candidate_406a46\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f94b4f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_819b1a\": 0.0, \"candidate_406a46\": 1.0, \"candidate_f94b4f\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:387f2e4c881490fa0f2b16a0:choice", "group_id": "fresh12:snli:387f2e4c881490fa0f2b16a0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Group of people wearing black shirts riding in an open top vehicle.\\nHypothesis: Froends ride in an open top vehicle together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e37ba2\": \"The hypothesis follows from the premise.\", \"candidate_e53819\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b0a1bb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e37ba2\": 0.0, \"candidate_e53819\": 1.0, \"candidate_b0a1bb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1d95a4781b42ef8eb2d75eaa:choice", "group_id": "fresh12:snli:1d95a4781b42ef8eb2d75eaa", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A dog jumps over a pile of wood.\\nHypothesis: A dog jumps over a pile of wood to chase a rabbit.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c50b14\": \"The hypothesis follows from the premise.\", \"candidate_559008\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_833c05\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c50b14\": 0.0, \"candidate_559008\": 1.0, \"candidate_833c05\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e20868a9ad6519e3cd712a40:choice", "group_id": "fresh12:snli:e20868a9ad6519e3cd712a40", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of people wait to cross the street in new york\\nHypothesis: Some of the people are wearing sunglasses.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6631f9\": \"The hypothesis follows from the premise.\", \"candidate_059186\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_cd5e1d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6631f9\": 0.0, \"candidate_059186\": 1.0, \"candidate_cd5e1d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6a358c206525c1ee904074e7:choice", "group_id": "fresh12:snli:6a358c206525c1ee904074e7", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The wind propels a sailing ship on a group of cruisers.\\nHypothesis: The ship nearly hits a cruiser.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_919cc5\": \"The hypothesis follows from the premise.\", \"candidate_d1502b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e88e0e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_919cc5\": 0.0, \"candidate_d1502b\": 1.0, \"candidate_e88e0e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b7efa935d5f94411248bf65d:choice", "group_id": "fresh12:snli:b7efa935d5f94411248bf65d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman, in red and white, and wearing glasses, sits in a room with other people.\\nHypothesis: A woman is on her phone sitting in a room.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5ea18c\": \"The hypothesis follows from the premise.\", \"candidate_848947\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8d8f5e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5ea18c\": 0.0, \"candidate_848947\": 1.0, \"candidate_8d8f5e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:48a3b784291939e5d3e88b80:choice", "group_id": "fresh12:snli:48a3b784291939e5d3e88b80", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a couple wearing bright outfits walking hand and hand on a sidewalk.\\nHypothesis: The outfits are orange.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2f6514\": \"The hypothesis follows from the premise.\", \"candidate_b5aa83\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5400fb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2f6514\": 0.0, \"candidate_b5aa83\": 1.0, \"candidate_5400fb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0d70b29c2a8b67502d2ea0f3:choice", "group_id": "fresh12:snli:0d70b29c2a8b67502d2ea0f3", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two old women sitting on curb next to flowers\\nHypothesis: Two old women sat near the scenery that they enjoyed\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_203368\": \"The hypothesis follows from the premise.\", \"candidate_9378b9\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e1f63d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_203368\": 0.0, \"candidate_9378b9\": 1.0, \"candidate_e1f63d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:68596a6e9eedb59387eb4119:choice", "group_id": "fresh12:snli:68596a6e9eedb59387eb4119", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of people having a conversation.\\nHypothesis: People are talking about politics.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_aaa160\": \"The hypothesis follows from the premise.\", \"candidate_9c6855\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_890a04\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_aaa160\": 0.0, \"candidate_9c6855\": 1.0, \"candidate_890a04\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ce3717c039b95904dc85d48c:choice", "group_id": "fresh12:snli:ce3717c039b95904dc85d48c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A snowboarder on a wide plain of snow\\nHypothesis: A snowmobile in a blizzard\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b5ca02\": \"The hypothesis follows from the premise.\", \"candidate_9aaf5f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f42478\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b5ca02\": 0.0, \"candidate_9aaf5f\": 1.0, \"candidate_f42478\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0586a466bcab4ed91dd692a5:choice", "group_id": "fresh12:snli:0586a466bcab4ed91dd692a5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two workers with leaf blowers blow orange leaves off patches of grass next to a wet sidewalk.\\nHypothesis: Two workers blowing leaves on a fall day\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a2200f\": \"The hypothesis follows from the premise.\", \"candidate_4ac578\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1763f2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a2200f\": 0.0, \"candidate_4ac578\": 1.0, \"candidate_1763f2\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:43281657016d72762da9508f:choice", "group_id": "fresh12:snli:43281657016d72762da9508f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A lady and a man in a hat watch baseball from the stands.\\nHypothesis: A woman and her husband are at a baseball game.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_051ffc\": \"The hypothesis follows from the premise.\", \"candidate_e023a9\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7f5fcc\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_051ffc\": 0.0, \"candidate_e023a9\": 1.0, \"candidate_7f5fcc\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:13cadb39ad2e43305bf48e54:choice", "group_id": "fresh12:snli:13cadb39ad2e43305bf48e54", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Person parasailing with a board in the ocean.\\nHypothesis: This person is looking at other people.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b5fe3b\": \"The hypothesis follows from the premise.\", \"candidate_ee536f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d313c2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b5fe3b\": 0.0, \"candidate_ee536f\": 1.0, \"candidate_d313c2\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1893df48539e5346fe9866fb:choice", "group_id": "fresh12:snli:1893df48539e5346fe9866fb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of people cleaning some kins of sea animal.\\nHypothesis: The people are at the beach.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1b5cdb\": \"The hypothesis follows from the premise.\", \"candidate_12e3ac\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_507fc2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1b5cdb\": 0.0, \"candidate_12e3ac\": 1.0, \"candidate_507fc2\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:27fc934f8dc8750885bbe045:choice", "group_id": "fresh12:snli:27fc934f8dc8750885bbe045", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A lady wearing a straw hat and gold hoop earrings is pointing a finger.\\nHypothesis: A woman points at the man she is talking about.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9f1db0\": \"The hypothesis follows from the premise.\", \"candidate_097ee7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0aa72d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9f1db0\": 0.0, \"candidate_097ee7\": 1.0, \"candidate_0aa72d\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:42afbe1631c64873acf6f0f2:choice", "group_id": "fresh12:snli:42afbe1631c64873acf6f0f2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A soft focus picture of a little girl wearing an orange sweater.\\nHypothesis: A girl is wearing an orange sweater with black stripes.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ce31fe\": \"The hypothesis follows from the premise.\", \"candidate_e8b20f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_69145e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ce31fe\": 0.0, \"candidate_e8b20f\": 1.0, \"candidate_69145e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bee29628ff9895aa1e1eadd1:choice", "group_id": "fresh12:snli:bee29628ff9895aa1e1eadd1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A younger man dressed in tribal attire.\\nHypothesis: A young man wheres tribal clothes for a ceremony.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2fa2ab\": \"The hypothesis follows from the premise.\", \"candidate_22b1f3\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_54499b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2fa2ab\": 0.0, \"candidate_22b1f3\": 1.0, \"candidate_54499b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:104d0b70a124e2aac0ea756a:choice", "group_id": "fresh12:snli:104d0b70a124e2aac0ea756a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: One man in baseball cap looks out while two men look down.\\nHypothesis: Men are gazing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3026e3\": \"The hypothesis follows from the premise.\", \"candidate_748fe4\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_dae376\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3026e3\": 0.0, \"candidate_748fe4\": 1.0, \"candidate_dae376\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:13b83eb7ee84e759a8e4243d:choice", "group_id": "fresh12:snli:13b83eb7ee84e759a8e4243d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two performers, one man playing a guitar, and one woman with a mic raised to her mouth are sitting on chairs performing.\\nHypothesis: The man and woman are doing karaoke together at a bar.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c8603c\": \"The hypothesis follows from the premise.\", \"candidate_50ab36\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7c1dfe\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c8603c\": 0.0, \"candidate_50ab36\": 1.0, \"candidate_7c1dfe\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:aed77fc1756ae4c30e1659a6:choice", "group_id": "fresh12:snli:aed77fc1756ae4c30e1659a6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A brown dog plays in a deep pile of snow.\\nHypothesis: A brown dog is outside and it's snowing\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_533a88\": \"The hypothesis follows from the premise.\", \"candidate_7e890c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5a3097\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_533a88\": 0.0, \"candidate_7e890c\": 1.0, \"candidate_5a3097\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:87d791880a7a6080968a64f5:choice", "group_id": "fresh12:snli:87d791880a7a6080968a64f5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Several cartoon characters are behind a young girl sitting in a highchair with a bucket and SpongeBob in front of her.\\nHypothesis: There's a gathering of cartoon characters.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_96c69f\": \"The hypothesis follows from the premise.\", \"candidate_726b01\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_be0e7f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_96c69f\": 0.0, \"candidate_726b01\": 1.0, \"candidate_be0e7f\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:222910b54839cb9bacfad91f:choice", "group_id": "fresh12:snli:222910b54839cb9bacfad91f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man is holding an accordion while leaning against a brown fence that overlooks some water.\\nHypothesis: An accordion player leans against his fence\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eca91e\": \"The hypothesis follows from the premise.\", \"candidate_a6ef55\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_02fc85\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eca91e\": 0.0, \"candidate_a6ef55\": 1.0, \"candidate_02fc85\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b5d2f6d1875825adbaf6b7de:choice", "group_id": "fresh12:snli:b5d2f6d1875825adbaf6b7de", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A little girl in a gray shirt smiles as she brushes her teeth with a yellow and orange toothbrush.\\nHypothesis: A girl fakes a smile to brush her teeth.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6f3a15\": \"The hypothesis follows from the premise.\", \"candidate_2bafc0\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9739ac\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6f3a15\": 0.0, \"candidate_2bafc0\": 1.0, \"candidate_9739ac\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:70686c56071d688355b94573:choice", "group_id": "fresh12:snli:70686c56071d688355b94573", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of female athletes are huddled together and excited.\\nHypothesis: They are about to plan a score.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3cceff\": \"The hypothesis follows from the premise.\", \"candidate_7a9b91\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3f48ef\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3cceff\": 0.0, \"candidate_7a9b91\": 1.0, \"candidate_3f48ef\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5172f44aa3099f6a0200ac67:choice", "group_id": "fresh12:snli:5172f44aa3099f6a0200ac67", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two police officers in yellow reflective jackets are standing in a shopping zone having a conversation.\\nHypothesis: Two police officers in yellow reflective jackets are standing in a busy shopping zone\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2f42e8\": \"The hypothesis follows from the premise.\", \"candidate_ec3028\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0a48c1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2f42e8\": 0.0, \"candidate_ec3028\": 1.0, \"candidate_0a48c1\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:132b8ecef42a556258cc93fb:choice", "group_id": "fresh12:snli:132b8ecef42a556258cc93fb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A statue at a museum that no seems to be looking at.\\nHypothesis: The statue is offensive and people are mad that it is on display.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_60cd77\": \"The hypothesis follows from the premise.\", \"candidate_a36bc7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_dd10f3\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_60cd77\": 0.0, \"candidate_a36bc7\": 1.0, \"candidate_dd10f3\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:87e23c98e54fe2d83d155931:choice", "group_id": "fresh12:snli:87e23c98e54fe2d83d155931", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man is riding a bike along a dirt trail.\\nHypothesis: The man is riding a dirt bike.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_da7473\": \"The hypothesis follows from the premise.\", \"candidate_ae41bb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b4b070\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_da7473\": 0.0, \"candidate_ae41bb\": 1.0, \"candidate_b4b070\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:22c9abf6e41c8c3293ccd8df:choice", "group_id": "fresh12:snli:22c9abf6e41c8c3293ccd8df", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A baby is eating aqua-colored cake very messily.\\nHypothesis: A baby eating cake celebrating it's birthday.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3380b7\": \"The hypothesis follows from the premise.\", \"candidate_fc5da1\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_be9f9b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3380b7\": 0.0, \"candidate_fc5da1\": 1.0, \"candidate_be9f9b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c25849a0b47352ea1d63032c:choice", "group_id": "fresh12:snli:c25849a0b47352ea1d63032c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The man in the black t-shirt is trying to throw something.\\nHypothesis: The man is throwing something to his son.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_df2943\": \"The hypothesis follows from the premise.\", \"candidate_fbb60c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_aa410e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_df2943\": 0.0, \"candidate_fbb60c\": 1.0, \"candidate_aa410e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:77922a5a752125ae2b11e4ba:choice", "group_id": "fresh12:snli:77922a5a752125ae2b11e4ba", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young man in an orange shirt, maneuvers a skateboard at the top of a ramp.\\nHypothesis: A person rides a skateboard well.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_739fd8\": \"The hypothesis follows from the premise.\", \"candidate_85de97\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c11a69\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_739fd8\": 0.0, \"candidate_85de97\": 1.0, \"candidate_c11a69\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:2a91a8ca2d642a44eae5bf70:choice", "group_id": "fresh12:snli:2a91a8ca2d642a44eae5bf70", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young girl is wearing a blue patterned swimsuit, a black and yellow swim cap and has blue swim goggles on her head.\\nHypothesis: Girl with swim cap and goggles prepares to jump in pool.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c0e518\": \"The hypothesis follows from the premise.\", \"candidate_6dcfae\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_179a62\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c0e518\": 0.0, \"candidate_6dcfae\": 1.0, \"candidate_179a62\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b0b6712be852f8b4d51762c5:choice", "group_id": "fresh12:snli:b0b6712be852f8b4d51762c5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A little girl in a pink dress crying.\\nHypothesis: A young girl is crying for more attention.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7904ee\": \"The hypothesis follows from the premise.\", \"candidate_98a7eb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d2402f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7904ee\": 0.0, \"candidate_98a7eb\": 1.0, \"candidate_d2402f\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0697f08683bf790513ee7812:choice", "group_id": "fresh12:snli:0697f08683bf790513ee7812", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A tall elder man is standing with his hand around another shorter elder man in the middle of the city\\nHypothesis: they are posing\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c75f44\": \"The hypothesis follows from the premise.\", \"candidate_1ad1fe\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c46377\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c75f44\": 0.0, \"candidate_1ad1fe\": 1.0, \"candidate_c46377\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b90c824b2b6cb6106dd07543:choice", "group_id": "fresh12:snli:b90c824b2b6cb6106dd07543", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Woman standing in a store with an item in hand.\\nHypothesis: A woman in a store is shopping.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d41a3a\": \"The hypothesis follows from the premise.\", \"candidate_fe5f0c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e689f5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d41a3a\": 0.0, \"candidate_fe5f0c\": 1.0, \"candidate_e689f5\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ffaf5cbbf21004e7677c13f0:choice", "group_id": "fresh12:snli:ffaf5cbbf21004e7677c13f0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man with a tan jacket with a full grocery bag is crossing the street.\\nHypothesis: The man just went shopping at the grocery store.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0b3d19\": \"The hypothesis follows from the premise.\", \"candidate_65a930\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e1416c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0b3d19\": 0.0, \"candidate_65a930\": 1.0, \"candidate_e1416c\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:22484db16c90c3b77440530e:choice", "group_id": "fresh12:snli:22484db16c90c3b77440530e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Man on phone in business suit walking down a busy street.\\nHypothesis: A man is talking to his wife on the phone while walking down the street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6d6a20\": \"The hypothesis follows from the premise.\", \"candidate_2a56a7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_95fec7\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6d6a20\": 0.0, \"candidate_2a56a7\": 1.0, \"candidate_95fec7\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:72f6f13f67ecb6a945d49d05:choice", "group_id": "fresh12:snli:72f6f13f67ecb6a945d49d05", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The people await to cross the street while walking the dog.\\nHypothesis: people are at a busy intersection\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_313ad9\": \"The hypothesis follows from the premise.\", \"candidate_6e6f5e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a5d377\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_313ad9\": 0.0, \"candidate_6e6f5e\": 1.0, \"candidate_a5d377\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cdc89b664dd559fed2fd2645:choice", "group_id": "fresh12:snli:cdc89b664dd559fed2fd2645", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A large crowd is gathered outside and one woman is yelling.\\nHypothesis: A protest is assembling.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b6527d\": \"The hypothesis follows from the premise.\", \"candidate_784605\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a1324a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b6527d\": 0.0, \"candidate_784605\": 1.0, \"candidate_a1324a\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bf6108363c965e76eabfe00f:choice", "group_id": "fresh12:snli:bf6108363c965e76eabfe00f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three people are sitting at a small table near a screen with the words, \\\"DEMOS, The Business of Care.\\\"\\nHypothesis: There are some people at a business meeting.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e3052a\": \"The hypothesis follows from the premise.\", \"candidate_b9b4dd\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9c21e5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e3052a\": 0.0, \"candidate_b9b4dd\": 1.0, \"candidate_9c21e5\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0d29180da9d210e1c4c62530:choice", "group_id": "fresh12:snli:0d29180da9d210e1c4c62530", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men with matching t-shirts fold a towel in the living room.\\nHypothesis: Two men with matching baseball team shirts are folding a towel.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_464db4\": \"The hypothesis follows from the premise.\", \"candidate_83793c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a6b659\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_464db4\": 0.0, \"candidate_83793c\": 1.0, \"candidate_a6b659\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5e7b1f306571f560bc14db67:choice", "group_id": "fresh12:snli:5e7b1f306571f560bc14db67", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three women sit at a table, taking notes from three identical magazines.\\nHypothesis: Three college women are studing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d2d5e9\": \"The hypothesis follows from the premise.\", \"candidate_44f58d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_093cdb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d2d5e9\": 0.0, \"candidate_44f58d\": 1.0, \"candidate_093cdb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:da74c4df9df22474d6bd23f6:choice", "group_id": "fresh12:snli:da74c4df9df22474d6bd23f6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man wearing protective gear grinds down a large metal object.\\nHypothesis: The man is using a grinder wearing red protective gear.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ca8a43\": \"The hypothesis follows from the premise.\", \"candidate_6134cf\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_16177e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ca8a43\": 0.0, \"candidate_6134cf\": 1.0, \"candidate_16177e\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1f7f76b4efa51999dbeca26c:choice", "group_id": "fresh12:snli:1f7f76b4efa51999dbeca26c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The blond boy is riding roller-coaster with his mouth wide open.\\nHypothesis: a little boy is riding a roller-coaster\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_42d0dc\": \"The hypothesis follows from the premise.\", \"candidate_820936\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4d14eb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_42d0dc\": 0.0, \"candidate_820936\": 1.0, \"candidate_4d14eb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f954cdaebaff0eececbf2f3d:choice", "group_id": "fresh12:snli:f954cdaebaff0eececbf2f3d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men stand in the center of a clearing with mountains in the background.\\nHypothesis: men at the center of a clearing waiting for the rest of the group\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_375a82\": \"The hypothesis follows from the premise.\", \"candidate_a55b50\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_cff398\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_375a82\": 0.0, \"candidate_a55b50\": 1.0, \"candidate_cff398\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4ec8ff2b2146ced4a115a85b:choice", "group_id": "fresh12:snli:4ec8ff2b2146ced4a115a85b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Men are playing rugby while covered in mud.\\nHypothesis: It rained recently.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8a25fc\": \"The hypothesis follows from the premise.\", \"candidate_e8a8fb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3dfeac\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8a25fc\": 0.0, \"candidate_e8a8fb\": 1.0, \"candidate_3dfeac\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:75b8e65ff93b57193f044ae2:choice", "group_id": "fresh12:snli:75b8e65ff93b57193f044ae2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy wearing a yellow shirt, khaki pants, and blue and yellow shoes is hitting a tennis ball off a wall with a tennis racket.\\nHypothesis: A child is getting ready for his junior athletic league tennis tournament.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1d7e71\": \"The hypothesis follows from the premise.\", \"candidate_d467e4\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7b06eb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1d7e71\": 0.0, \"candidate_d467e4\": 1.0, \"candidate_7b06eb\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4161b428cf9d0e5a5ebee683:choice", "group_id": "fresh12:snli:4161b428cf9d0e5a5ebee683", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of young cheerleaders in pink and white costumes perform on a stage.\\nHypothesis: They are holding pom poms.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_65c095\": \"The hypothesis follows from the premise.\", \"candidate_e325fc\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_cdfbc8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_65c095\": 0.0, \"candidate_e325fc\": 1.0, \"candidate_cdfbc8\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b6894a1094b775321824eb73:choice", "group_id": "fresh12:snli:b6894a1094b775321824eb73", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A swan landing in a lake.\\nHypothesis: A bird is looking for food on a lake.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_62c0e4\": \"The hypothesis follows from the premise.\", \"candidate_df2704\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e81b6b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_62c0e4\": 0.0, \"candidate_df2704\": 1.0, \"candidate_e81b6b\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9877ea5a992bdedc0141b06e:choice", "group_id": "fresh12:snli:9877ea5a992bdedc0141b06e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in a brown shirt is kneeling in the street.\\nHypothesis: A woman is protesting in the street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fda0d7\": \"The hypothesis follows from the premise.\", \"candidate_ba84bd\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_032340\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fda0d7\": 0.0, \"candidate_ba84bd\": 1.0, \"candidate_032340\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:51484bee38b7c6ef253c5b71:choice", "group_id": "fresh12:snli:51484bee38b7c6ef253c5b71", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man playing a large stringed instrument on the street.\\nHypothesis: There is a man entertaining outdoors.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_60e729\": \"The hypothesis follows from the premise.\", \"candidate_5a46cd\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7d8269\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_60e729\": 0.0, \"candidate_5a46cd\": 1.0, \"candidate_7d8269\": 0.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:845e9487459bd3ec0daca151:choice", "group_id": "fresh12:snli:845e9487459bd3ec0daca151", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The older woman smiles as she holds up a hammer.\\nHypothesis: A man is kicking some rocks.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d79cac\": \"The hypothesis follows from the premise.\", \"candidate_ce0a2c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b7f76c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d79cac\": 0.0, \"candidate_ce0a2c\": 0.0, \"candidate_b7f76c\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e86aee4e65ee653ee5b43ecf:choice", "group_id": "fresh12:snli:e86aee4e65ee653ee5b43ecf", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: People are shopping around Christmas time for decorations.\\nHypothesis: People are buying fireworks at a dealer.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7bcf84\": \"The hypothesis follows from the premise.\", \"candidate_2a8643\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2801cb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7bcf84\": 0.0, \"candidate_2a8643\": 0.0, \"candidate_2801cb\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:50b66e162bf3dea6342fa99a:choice", "group_id": "fresh12:snli:50b66e162bf3dea6342fa99a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man with a latex glove and black t-shit is standing in front of a Chevy van.\\nHypothesis: a man in his van because he caught his victim\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1442eb\": \"The hypothesis follows from the premise.\", \"candidate_1a076a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d797f5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1442eb\": 0.0, \"candidate_1a076a\": 0.0, \"candidate_d797f5\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:28081bd1ff41042df68e8b6a:choice", "group_id": "fresh12:snli:28081bd1ff41042df68e8b6a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man and a Woman cutting dough.\\nHypothesis: A man and woman eating vegetables.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a23fe1\": \"The hypothesis follows from the premise.\", \"candidate_1f0a27\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9718df\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a23fe1\": 0.0, \"candidate_1f0a27\": 0.0, \"candidate_9718df\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:70a1256126fedae42f62e23f:choice", "group_id": "fresh12:snli:70a1256126fedae42f62e23f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Four young female gymnasts dressed in black and blue leotards are preparing for their next routine.\\nHypothesis: Band performs onstage to an empty room.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f35392\": \"The hypothesis follows from the premise.\", \"candidate_9f05a9\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_6eeded\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f35392\": 0.0, \"candidate_9f05a9\": 0.0, \"candidate_6eeded\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:22449f59042fdf28b1f941ef:choice", "group_id": "fresh12:snli:22449f59042fdf28b1f941ef", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A white dog with brown spots running on a dirt path\\nHypothesis: A black dog walks on a concrete path.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_00924f\": \"The hypothesis follows from the premise.\", \"candidate_9136d0\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_6c208c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_00924f\": 0.0, \"candidate_9136d0\": 0.0, \"candidate_6c208c\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:464e56181758316f7bcbd9db:choice", "group_id": "fresh12:snli:464e56181758316f7bcbd9db", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A shirtless man with his fishing boat.\\nHypothesis: A man is sleeping at home.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3d2e51\": \"The hypothesis follows from the premise.\", \"candidate_cf4ea4\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_34cd8b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3d2e51\": 0.0, \"candidate_cf4ea4\": 0.0, \"candidate_34cd8b\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b6aeca5d688f39f8514f06f3:choice", "group_id": "fresh12:snli:b6aeca5d688f39f8514f06f3", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A cheerful boy rides a makeshift sled down a snowy hill.\\nHypothesis: Don't stock up and have fun with the sled\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1fdc08\": \"The hypothesis follows from the premise.\", \"candidate_69796c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_de07c1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1fdc08\": 0.0, \"candidate_69796c\": 0.0, \"candidate_de07c1\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:59c3655f84f96445e6069958:choice", "group_id": "fresh12:snli:59c3655f84f96445e6069958", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman is taking a picture on the shoulders of another woman in cold weather.\\nHypothesis: A man is running errands on his bike.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_394d80\": \"The hypothesis follows from the premise.\", \"candidate_b9aa41\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_96b8c9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_394d80\": 0.0, \"candidate_b9aa41\": 0.0, \"candidate_96b8c9\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f17e9eb8e24be0a965a112d6:choice", "group_id": "fresh12:snli:f17e9eb8e24be0a965a112d6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men are drumming, one is standing and one is sitting down.\\nHypothesis: Two women are playing the flute.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_de475c\": \"The hypothesis follows from the premise.\", \"candidate_def170\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a86fa6\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_de475c\": 0.0, \"candidate_def170\": 0.0, \"candidate_a86fa6\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5d477c5af4f9e3bb70408fda:choice", "group_id": "fresh12:snli:5d477c5af4f9e3bb70408fda", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man is working on a Earthen structure with a pickaxe.\\nHypothesis: The man is in an office building.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9ad98c\": \"The hypothesis follows from the premise.\", \"candidate_69d05e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_14daba\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9ad98c\": 0.0, \"candidate_69d05e\": 0.0, \"candidate_14daba\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:89b1247e749cc8f7eab0a4a8:choice", "group_id": "fresh12:snli:89b1247e749cc8f7eab0a4a8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young man standing on one foot in colorful bowling shoes after throwing the ball.\\nHypothesis: A young man is standing on both feet barefooted.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6003da\": \"The hypothesis follows from the premise.\", \"candidate_c470ee\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d8aa94\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6003da\": 0.0, \"candidate_c470ee\": 0.0, \"candidate_d8aa94\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ef201f11d5d6328b10fd87e8:choice", "group_id": "fresh12:snli:ef201f11d5d6328b10fd87e8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in a red dress is walking on the sidewalk.\\nHypothesis: The woman is running through the woods.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e8501f\": \"The hypothesis follows from the premise.\", \"candidate_f87678\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_881b89\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e8501f\": 0.0, \"candidate_f87678\": 0.0, \"candidate_881b89\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:76b55636f3660a08ab6d1c28:choice", "group_id": "fresh12:snli:76b55636f3660a08ab6d1c28", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man and a woman are standing at the side of an inflatable climbing structure.\\nHypothesis: A man and a woman are seated\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_be5e11\": \"The hypothesis follows from the premise.\", \"candidate_751fcb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0c87af\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_be5e11\": 0.0, \"candidate_751fcb\": 0.0, \"candidate_0c87af\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:126ba77865ec65661912b62c:choice", "group_id": "fresh12:snli:126ba77865ec65661912b62c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two young musicians play from a stage.\\nHypothesis: The stage is empty\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_517c18\": \"The hypothesis follows from the premise.\", \"candidate_64f4b5\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_485ab6\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_517c18\": 0.0, \"candidate_64f4b5\": 0.0, \"candidate_485ab6\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9ca4b9ccde25c1c2eb73bf10:choice", "group_id": "fresh12:snli:9ca4b9ccde25c1c2eb73bf10", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of men playing rugby on the sand.\\nHypothesis: A group of woman play volleyball in a court.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bb458e\": \"The hypothesis follows from the premise.\", \"candidate_d975d8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_439ccd\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bb458e\": 0.0, \"candidate_d975d8\": 0.0, \"candidate_439ccd\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:00a76b0c0545ca97134b082d:choice", "group_id": "fresh12:snli:00a76b0c0545ca97134b082d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Young children waving American flags for a soldier\\nHypothesis: Children are burning flags in protest.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a0420c\": \"The hypothesis follows from the premise.\", \"candidate_2d7bc6\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7a9eeb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a0420c\": 0.0, \"candidate_2d7bc6\": 0.0, \"candidate_7a9eeb\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5bf3ba2494cfe48ecab2f4cb:choice", "group_id": "fresh12:snli:5bf3ba2494cfe48ecab2f4cb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An older women looking off into the distance with three older women walking towards her from behind.\\nHypothesis: Four older women are sitting in a hot tub\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_50ee02\": \"The hypothesis follows from the premise.\", \"candidate_09821c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3e484e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_50ee02\": 0.0, \"candidate_09821c\": 0.0, \"candidate_3e484e\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:70dccc6af9c430872e7e1c6b:choice", "group_id": "fresh12:snli:70dccc6af9c430872e7e1c6b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Pedestrians use the crosswalks as traffic awaits their turn to drive through an intersection.\\nHypothesis: Traffic drives through the crosswalk.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7ea8d7\": \"The hypothesis follows from the premise.\", \"candidate_a98238\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f0547d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7ea8d7\": 0.0, \"candidate_a98238\": 0.0, \"candidate_f0547d\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6de8e94d3826cd1d4f71fa23:choice", "group_id": "fresh12:snli:6de8e94d3826cd1d4f71fa23", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An elderly man with glasses, a red ball cap and backpack is standing in front of an open doorway.\\nHypothesis: the man is flying a plane\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_92feea\": \"The hypothesis follows from the premise.\", \"candidate_60c9ca\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f29384\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_92feea\": 0.0, \"candidate_60c9ca\": 0.0, \"candidate_f29384\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6b26f5e9e6e18ee751dfd6df:choice", "group_id": "fresh12:snli:6b26f5e9e6e18ee751dfd6df", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A person holds a bunch of brush while walking down a stone path.\\nHypothesis: The person is wading through water.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3efe4c\": \"The hypothesis follows from the premise.\", \"candidate_64142a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_dc85b9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3efe4c\": 0.0, \"candidate_64142a\": 0.0, \"candidate_dc85b9\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7c3112a2e391bbe8455d8e98:choice", "group_id": "fresh12:snli:7c3112a2e391bbe8455d8e98", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men entering a brick home during the snow.\\nHypothesis: It's a bright summer day when the men enter the home\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ec3263\": \"The hypothesis follows from the premise.\", \"candidate_1f899f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0f7a8b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ec3263\": 0.0, \"candidate_1f899f\": 0.0, \"candidate_0f7a8b\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:548736e91dd3ca1d85080d80:choice", "group_id": "fresh12:snli:548736e91dd3ca1d85080d80", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: There are a group of track runners bracing to take off.\\nHypothesis: The  track runners are sitting on an airplane\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d86ec9\": \"The hypothesis follows from the premise.\", \"candidate_aa8ad5\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_90a593\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d86ec9\": 0.0, \"candidate_aa8ad5\": 0.0, \"candidate_90a593\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f90ea2c720b9501be8300a80:choice", "group_id": "fresh12:snli:f90ea2c720b9501be8300a80", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three girls jumping in front of a house's flower garden.\\nHypothesis: The girls are playing on a street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d406c8\": \"The hypothesis follows from the premise.\", \"candidate_061570\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_01ce4f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d406c8\": 0.0, \"candidate_061570\": 0.0, \"candidate_01ce4f\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:80cf3c635ecc965faf74506b:choice", "group_id": "fresh12:snli:80cf3c635ecc965faf74506b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Small dog in costume stands on hind legs to reach dangling flowers.\\nHypothesis: A dog is trying to eat a dangling kitten on a tree.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c4c516\": \"The hypothesis follows from the premise.\", \"candidate_5c9783\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_82f9d9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c4c516\": 0.0, \"candidate_5c9783\": 0.0, \"candidate_82f9d9\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ad63a13d71c838511f7b70df:choice", "group_id": "fresh12:snli:ad63a13d71c838511f7b70df", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A bearded man with a backpack on his back holding a tripod in front of a rocky creek and bridge.\\nHypothesis: The man was sitting indoors watching TV.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8c3679\": \"The hypothesis follows from the premise.\", \"candidate_3df11b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b82ab2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8c3679\": 0.0, \"candidate_3df11b\": 0.0, \"candidate_b82ab2\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:788dfa8b6d416c7ab829c130:choice", "group_id": "fresh12:snli:788dfa8b6d416c7ab829c130", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two racers race on motorcycles.\\nHypothesis: Two people are racing cars.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b58d07\": \"The hypothesis follows from the premise.\", \"candidate_defaad\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d63b46\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b58d07\": 0.0, \"candidate_defaad\": 0.0, \"candidate_d63b46\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:8ecc2fd7e1eb0dbabde34a01:choice", "group_id": "fresh12:snli:8ecc2fd7e1eb0dbabde34a01", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two hussars sit perched on horses, dressed in extravagant ceremonial wear, each holding a sabre in their right hand, reigns to the horse in their left.\\nHypothesis: The men are driving their motorcycles.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_094eaa\": \"The hypothesis follows from the premise.\", \"candidate_86875f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_17e575\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_094eaa\": 0.0, \"candidate_86875f\": 0.0, \"candidate_17e575\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:240f85663bd75c6a1f241535:choice", "group_id": "fresh12:snli:240f85663bd75c6a1f241535", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young girl in a red top is running through a grassy area with large buildings in the background.\\nHypothesis: The girl is driving in the city.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fbacb9\": \"The hypothesis follows from the premise.\", \"candidate_b856ab\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_533140\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fbacb9\": 0.0, \"candidate_b856ab\": 0.0, \"candidate_533140\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:04fd19cdc85660b54e8acecc:choice", "group_id": "fresh12:snli:04fd19cdc85660b54e8acecc", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young girl in a white t-shirt and a blue skirt wearing a light colored hat stands in a river next to a big rock.\\nHypothesis: The girl is swimming in the river.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5e34c1\": \"The hypothesis follows from the premise.\", \"candidate_706a51\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_005476\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5e34c1\": 0.0, \"candidate_706a51\": 0.0, \"candidate_005476\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:98398b8b31003870298ede69:choice", "group_id": "fresh12:snli:98398b8b31003870298ede69", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a black tank top wearing a red plaid hat\\nHypothesis: A man wearing football pads.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_76b5cc\": \"The hypothesis follows from the premise.\", \"candidate_892c93\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b12fe5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_76b5cc\": 0.0, \"candidate_892c93\": 0.0, \"candidate_b12fe5\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9e2ed2a34596bb46f41544da:choice", "group_id": "fresh12:snli:9e2ed2a34596bb46f41544da", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men with black glasses are holding food, one appears to be eating.\\nHypothesis: A group of women is drinking wine.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_37d747\": \"The hypothesis follows from the premise.\", \"candidate_1eafec\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_fdd6e0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_37d747\": 0.0, \"candidate_1eafec\": 0.0, \"candidate_fdd6e0\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1ff44b004c99ed48e3093846:choice", "group_id": "fresh12:snli:1ff44b004c99ed48e3093846", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young japanese couple holding an umbrella walking down a street.\\nHypothesis: A young Asian couple are swimming in the pool.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c82808\": \"The hypothesis follows from the premise.\", \"candidate_0bc8d8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_295493\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c82808\": 0.0, \"candidate_0bc8d8\": 0.0, \"candidate_295493\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7bc50c9a6631bfe339c2026d:choice", "group_id": "fresh12:snli:7bc50c9a6631bfe339c2026d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: several brides wearing white dresses are holding red roses\\nHypothesis: There is one bride.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cf30d2\": \"The hypothesis follows from the premise.\", \"candidate_1f52d0\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_26d39e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cf30d2\": 0.0, \"candidate_1f52d0\": 0.0, \"candidate_26d39e\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4eca7960b6f6d229a154446f:choice", "group_id": "fresh12:snli:4eca7960b6f6d229a154446f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a woman in a brown dress looking at papers in front of a class.\\nHypothesis: A woman in a brown dress is a murderous clown at a birthday party.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a16f67\": \"The hypothesis follows from the premise.\", \"candidate_bb8a40\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_144203\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a16f67\": 0.0, \"candidate_bb8a40\": 0.0, \"candidate_144203\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:96efece8dad38a96eb10f782:choice", "group_id": "fresh12:snli:96efece8dad38a96eb10f782", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man is on a skateboard jumping from a short ledge onto the sidewalk.\\nHypothesis: The man is walking with his skateboard on the street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9205ab\": \"The hypothesis follows from the premise.\", \"candidate_471794\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a4c79b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9205ab\": 0.0, \"candidate_471794\": 0.0, \"candidate_a4c79b\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c7d2e40e9c8314d94ff36779:choice", "group_id": "fresh12:snli:c7d2e40e9c8314d94ff36779", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A local street in an Arabic country where there is a market selling Coca-Cola.\\nHypothesis: A snowy mountain in the Rockies.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_55bcba\": \"The hypothesis follows from the premise.\", \"candidate_a84cef\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_eeccea\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_55bcba\": 0.0, \"candidate_a84cef\": 0.0, \"candidate_eeccea\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bd3aabff0f679b86026b9bb0:choice", "group_id": "fresh12:snli:bd3aabff0f679b86026b9bb0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of cyclists at a race with their arms around each other.\\nHypothesis: A group of gymnasts performing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_976c94\": \"The hypothesis follows from the premise.\", \"candidate_9da989\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9b7970\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_976c94\": 0.0, \"candidate_9da989\": 0.0, \"candidate_9b7970\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4bb863023ea06440ba21644d:choice", "group_id": "fresh12:snli:4bb863023ea06440ba21644d", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of motorcycles wearing black.\\nHypothesis: Some motorcycles wearing white.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a5b845\": \"The hypothesis follows from the premise.\", \"candidate_be79f7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_bb1996\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a5b845\": 0.0, \"candidate_be79f7\": 0.0, \"candidate_bb1996\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:69f1203e4440bfa94cdacff9:choice", "group_id": "fresh12:snli:69f1203e4440bfa94cdacff9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three black people, two men and one woman, in the middle of a small, rural street, are piling up items on top of a taxi cab with a net dangling down on top of the windshield, while two children and two other people are nearby.\\nHypothesis: The taxi has nothing on it\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0afa06\": \"The hypothesis follows from the premise.\", \"candidate_fdcf3e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_86501e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0afa06\": 0.0, \"candidate_fdcf3e\": 0.0, \"candidate_86501e\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b887b8a8644d18e2251dd8ab:choice", "group_id": "fresh12:snli:b887b8a8644d18e2251dd8ab", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Men playing basketball, the guy with the ball is attempting a layup.\\nHypothesis: Men are playing basketball, the guy with the ball is shooting a free throw.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d1f0b5\": \"The hypothesis follows from the premise.\", \"candidate_d571be\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d1a591\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d1f0b5\": 0.0, \"candidate_d571be\": 0.0, \"candidate_d1a591\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:27fbeb9d373d9973b2fd847c:choice", "group_id": "fresh12:snli:27fbeb9d373d9973b2fd847c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of Chinese people sitting and laying down together on what appears to be a front porch.\\nHypothesis: A group of African American people are eating dinner at KFC.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3f6ae9\": \"The hypothesis follows from the premise.\", \"candidate_b526fb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_922adb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3f6ae9\": 0.0, \"candidate_b526fb\": 0.0, \"candidate_922adb\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:db1276f950801849cfe4c2c6:choice", "group_id": "fresh12:snli:db1276f950801849cfe4c2c6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman fell in the water after riding a wakeboard.\\nHypothesis: A woman lays on a pool float.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_be4dd9\": \"The hypothesis follows from the premise.\", \"candidate_4fcb78\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_cc7700\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_be4dd9\": 0.0, \"candidate_4fcb78\": 0.0, \"candidate_cc7700\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:fd991d5e4246c9e1089e36ef:choice", "group_id": "fresh12:snli:fd991d5e4246c9e1089e36ef", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The rhythmic gymnast completes her floor exercise at the competition.\\nHypothesis: The gymnast cannot finish her exercise.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b932d1\": \"The hypothesis follows from the premise.\", \"candidate_3b4365\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7cb9e8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b932d1\": 0.0, \"candidate_3b4365\": 0.0, \"candidate_7cb9e8\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:dd2b89dadc93db43598d998b:choice", "group_id": "fresh12:snli:dd2b89dadc93db43598d998b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy in a blue jacket rides a skateboard down the street.\\nHypothesis: A blue jacketed boy is standing in football field.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_06fb96\": \"The hypothesis follows from the premise.\", \"candidate_96cafe\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4d309c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_06fb96\": 0.0, \"candidate_96cafe\": 0.0, \"candidate_4d309c\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:55f5d19940ca4cc56937d2c1:choice", "group_id": "fresh12:snli:55f5d19940ca4cc56937d2c1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two woman, one wearing a blue striped shirt and the other wearing a flowered shirt, are riding horses.\\nHypothesis: Women are outside at a riding competition where they watch men compete.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_481d68\": \"The hypothesis follows from the premise.\", \"candidate_565f37\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_bba0b0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_481d68\": 0.0, \"candidate_565f37\": 0.0, \"candidate_bba0b0\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:062778cfa2c361e22f093aa8:choice", "group_id": "fresh12:snli:062778cfa2c361e22f093aa8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Soldiers are holding guns and peeking around a building.\\nHypothesis: Soldiers are holding guns in a large featureless desert.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d45690\": \"The hypothesis follows from the premise.\", \"candidate_3d2c63\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ac5743\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d45690\": 0.0, \"candidate_3d2c63\": 0.0, \"candidate_ac5743\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d687fb35b0412fe82b86685e:choice", "group_id": "fresh12:snli:d687fb35b0412fe82b86685e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Four men and a young boy are at a table set up for wood crafts.\\nHypothesis: Four men and a young boy are at the lake swimming.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_321e50\": \"The hypothesis follows from the premise.\", \"candidate_4d473a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f58dc2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_321e50\": 0.0, \"candidate_4d473a\": 0.0, \"candidate_f58dc2\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a1b9ad7c9e8874b618564265:choice", "group_id": "fresh12:snli:a1b9ad7c9e8874b618564265", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A bicycle rider wearing racing gear pedals a yellow bike past the wire fence at the edge of a field, with a stand of trees in the background.\\nHypothesis: A scooter rider wearing casual clothes races past a building.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9edd6e\": \"The hypothesis follows from the premise.\", \"candidate_8a6a48\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e07205\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9edd6e\": 0.0, \"candidate_8a6a48\": 0.0, \"candidate_e07205\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5a297f3ee8b64249f62eedec:choice", "group_id": "fresh12:snli:5a297f3ee8b64249f62eedec", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two women in bathing suits play on the beach.\\nHypothesis: Two girls are helping their mom with the laundry.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_803a5a\": \"The hypothesis follows from the premise.\", \"candidate_10ca16\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c90f76\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_803a5a\": 0.0, \"candidate_10ca16\": 0.0, \"candidate_c90f76\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cea7f46275a2d11385bda120:choice", "group_id": "fresh12:snli:cea7f46275a2d11385bda120", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three women talk to each other next to another woman sitting on a motorcycle.\\nHypothesis: A group of men are indoors watching football while it snows outside.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f31ddf\": \"The hypothesis follows from the premise.\", \"candidate_8afb79\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a99117\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f31ddf\": 0.0, \"candidate_8afb79\": 0.0, \"candidate_a99117\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ce6ff272d48fcf2c87b1998e:choice", "group_id": "fresh12:snli:ce6ff272d48fcf2c87b1998e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men working on the roof of an apartment building with a nice looking skyline behind them.\\nHypothesis: Two men not working on the roof of an apartment building\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_00e8b6\": \"The hypothesis follows from the premise.\", \"candidate_ea03be\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b4e7ce\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_00e8b6\": 0.0, \"candidate_ea03be\": 0.0, \"candidate_b4e7ce\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ba83f83056fa6358f179e7e2:choice", "group_id": "fresh12:snli:ba83f83056fa6358f179e7e2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A policeman on a motorcycle from the rear.\\nHypothesis: An officer watching a jail cell.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e0ec53\": \"The hypothesis follows from the premise.\", \"candidate_ef5f69\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2ea32d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e0ec53\": 0.0, \"candidate_ef5f69\": 0.0, \"candidate_2ea32d\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:59bd60b0247ea79639e8f59a:choice", "group_id": "fresh12:snli:59bd60b0247ea79639e8f59a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a man playing guitar and signing next to a woman in a dress playing an instrument\\nHypothesis: A dog is playing drums.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bd874a\": \"The hypothesis follows from the premise.\", \"candidate_a8dd87\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_88a03f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bd874a\": 0.0, \"candidate_a8dd87\": 0.0, \"candidate_88a03f\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:fbe468b611f06ee0344a5ed1:choice", "group_id": "fresh12:snli:fbe468b611f06ee0344a5ed1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: There's a red bus making a left turn into a traffic circle that has a sprinkler system.\\nHypothesis: A red bus at a bus stop\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e20da5\": \"The hypothesis follows from the premise.\", \"candidate_39591c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_809e8f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e20da5\": 0.0, \"candidate_39591c\": 0.0, \"candidate_809e8f\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ef6da9818d21595ad953e60c:choice", "group_id": "fresh12:snli:ef6da9818d21595ad953e60c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A yellow dog is running in a field near a mountain.\\nHypothesis: A yellow dog is going to the vet.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d92eda\": \"The hypothesis follows from the premise.\", \"candidate_19c1e3\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_35ab90\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d92eda\": 0.0, \"candidate_19c1e3\": 0.0, \"candidate_35ab90\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1effa374f64ec9ffd677f54c:choice", "group_id": "fresh12:snli:1effa374f64ec9ffd677f54c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An old woman wearing a ski hat and glasses is sitting on a blue chair.\\nHypothesis: Nobody is sitting.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a974bc\": \"The hypothesis follows from the premise.\", \"candidate_fedb27\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b49057\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a974bc\": 0.0, \"candidate_fedb27\": 0.0, \"candidate_b49057\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7b129d6fed8871ae827e83f1:choice", "group_id": "fresh12:snli:7b129d6fed8871ae827e83f1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A dog is walking down steps.\\nHypothesis: A dog is laying on the couch.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_231090\": \"The hypothesis follows from the premise.\", \"candidate_15042e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8b3d27\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_231090\": 0.0, \"candidate_15042e\": 0.0, \"candidate_8b3d27\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d0d15fc7fb35d12bbf4100e1:choice", "group_id": "fresh12:snli:d0d15fc7fb35d12bbf4100e1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man sits writing at a table set up on a sidewalk.\\nHypothesis: A woman stands on a street corner.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9bb2b6\": \"The hypothesis follows from the premise.\", \"candidate_286267\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_36b88f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9bb2b6\": 0.0, \"candidate_286267\": 0.0, \"candidate_36b88f\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6fb27958a8c34798f3d9e8a8:choice", "group_id": "fresh12:snli:6fb27958a8c34798f3d9e8a8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A music concerto is happening on stage, with the lights on spot, while the audience listens in the dark.\\nHypothesis: The people are at a play.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_17cb2d\": \"The hypothesis follows from the premise.\", \"candidate_51b249\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9bbb85\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_17cb2d\": 0.0, \"candidate_51b249\": 0.0, \"candidate_9bbb85\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:844b2cd7f41f5e71124823d1:choice", "group_id": "fresh12:snli:844b2cd7f41f5e71124823d1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of four children pose on a muddy beach, smiling and making faces.\\nHypothesis: A man stirs the stew.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b381c5\": \"The hypothesis follows from the premise.\", \"candidate_616e6a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5977a8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b381c5\": 0.0, \"candidate_616e6a\": 0.0, \"candidate_5977a8\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5496087630379bb2b8dc58ae:choice", "group_id": "fresh12:snli:5496087630379bb2b8dc58ae", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A little girl with pointy pigtails is sitting in the sand on the beach.\\nHypothesis: The girl runs through the grass outside her school.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_90f190\": \"The hypothesis follows from the premise.\", \"candidate_c1f408\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_cb8cee\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_90f190\": 0.0, \"candidate_c1f408\": 0.0, \"candidate_cb8cee\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:458aca3a46aa08f65cbf4cf8:choice", "group_id": "fresh12:snli:458aca3a46aa08f65cbf4cf8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young boy gets his photo taken with Mr. and Mrs. Clause.\\nHypothesis: A young girl gets his photo taken with Mr. and Mrs. Clause.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9cc7ad\": \"The hypothesis follows from the premise.\", \"candidate_b78cc8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c02232\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9cc7ad\": 0.0, \"candidate_b78cc8\": 0.0, \"candidate_c02232\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:fb9a70d5b7aa95cacbe3adac:choice", "group_id": "fresh12:snli:fb9a70d5b7aa95cacbe3adac", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: People protesting age discrimination in an organized street rally.\\nHypothesis: People are inside watching tv.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_88f248\": \"The hypothesis follows from the premise.\", \"candidate_8fc37b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ac4490\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_88f248\": 0.0, \"candidate_8fc37b\": 0.0, \"candidate_ac4490\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c7ee93444a775e3217eb208e:choice", "group_id": "fresh12:snli:c7ee93444a775e3217eb208e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Construction workers use cables to hang from the sides of buildings.\\nHypothesis: Construction workers are having fun smashing windows.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_347002\": \"The hypothesis follows from the premise.\", \"candidate_3f90a4\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9097d8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_347002\": 0.0, \"candidate_3f90a4\": 0.0, \"candidate_9097d8\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9c16beddeff0a1918af4dcc2:choice", "group_id": "fresh12:snli:9c16beddeff0a1918af4dcc2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A family with young children sits down outside.\\nHypothesis: A father is walking the dog with his children.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6db2e8\": \"The hypothesis follows from the premise.\", \"candidate_ae1f7b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_428d62\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6db2e8\": 0.0, \"candidate_ae1f7b\": 0.0, \"candidate_428d62\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6ce68f6a998970a8d02ed191:choice", "group_id": "fresh12:snli:6ce68f6a998970a8d02ed191", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in an American military uniform sits at a table and writes the words \\\"sad,\\\" \\\"depressed,\\\" and \\\"hatred\\\" on a large sheet of white paper.\\nHypothesis: A woman is diving off a mountain.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fc0cb7\": \"The hypothesis follows from the premise.\", \"candidate_a8c163\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4b7879\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fc0cb7\": 0.0, \"candidate_a8c163\": 0.0, \"candidate_4b7879\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:786407555170aeef0290e427:choice", "group_id": "fresh12:snli:786407555170aeef0290e427", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A squirrel stands on its hind legs in the grass at the edge of a road.\\nHypothesis: A squirrel is sleeping in the tree.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3f56c1\": \"The hypothesis follows from the premise.\", \"candidate_556d8d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5a5d24\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3f56c1\": 0.0, \"candidate_556d8d\": 0.0, \"candidate_5a5d24\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7860b0ad76b5149dc9880f3a:choice", "group_id": "fresh12:snli:7860b0ad76b5149dc9880f3a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A runner in yellow has a convoy of motorcycles following behind him on a highway as bystanders watch.\\nHypothesis: A speedwalker next to cars\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e087fe\": \"The hypothesis follows from the premise.\", \"candidate_e9ff71\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_66c4f1\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e087fe\": 0.0, \"candidate_e9ff71\": 0.0, \"candidate_66c4f1\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:14951444655db4b057709029:choice", "group_id": "fresh12:snli:14951444655db4b057709029", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A black-haired woman with her hair in a high bun is having her head shaved by another woman.\\nHypothesis: Everybody is bald\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_899bb3\": \"The hypothesis follows from the premise.\", \"candidate_b6571d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_66432f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_899bb3\": 0.0, \"candidate_b6571d\": 0.0, \"candidate_66432f\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9cfec7e5358d4f520c287b2b:choice", "group_id": "fresh12:snli:9cfec7e5358d4f520c287b2b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a brown dog coming out of an orange and yellow tunnel\\nHypothesis: A brown dog dancing out of a orange and yellow tunnel.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f3ca19\": \"The hypothesis follows from the premise.\", \"candidate_9132a9\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_24a5b6\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f3ca19\": 0.0, \"candidate_9132a9\": 0.0, \"candidate_24a5b6\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:3137ece9b4340f3d874113dc:choice", "group_id": "fresh12:snli:3137ece9b4340f3d874113dc", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The African American man protests against unlawful sex.\\nHypothesis: The man sings and dances in the park.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a0ce4b\": \"The hypothesis follows from the premise.\", \"candidate_c8d4ad\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_803d10\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a0ce4b\": 0.0, \"candidate_c8d4ad\": 0.0, \"candidate_803d10\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:722a8e54274b89df17d42337:choice", "group_id": "fresh12:snli:722a8e54274b89df17d42337", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man wearing black pants rides along a concrete ledge on rollerblades.\\nHypothesis: The roller blade wheels are square.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5843fa\": \"The hypothesis follows from the premise.\", \"candidate_cb07e9\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b1e407\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5843fa\": 0.0, \"candidate_cb07e9\": 0.0, \"candidate_b1e407\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e19b918f81100bd716c5119a:choice", "group_id": "fresh12:snli:e19b918f81100bd716c5119a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men on opposing teams are playing soccer in a field.\\nHypothesis: A man crochets.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3e8453\": \"The hypothesis follows from the premise.\", \"candidate_4fe323\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9d4d99\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3e8453\": 0.0, \"candidate_4fe323\": 0.0, \"candidate_9d4d99\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a7a086471f641edababf9de5:choice", "group_id": "fresh12:snli:a7a086471f641edababf9de5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young girl is standing in a kitchen holding a green bib.\\nHypothesis: A boy is playing with a firetruck.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_405575\": \"The hypothesis follows from the premise.\", \"candidate_46315d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_430bbd\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_405575\": 0.0, \"candidate_46315d\": 0.0, \"candidate_430bbd\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4a12df5552142083098147b0:choice", "group_id": "fresh12:snli:4a12df5552142083098147b0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy is being splashed in a swimming pool.\\nHypothesis: A boy is splashing others at the beach.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_409655\": \"The hypothesis follows from the premise.\", \"candidate_9cd644\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a46ad7\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_409655\": 0.0, \"candidate_9cd644\": 0.0, \"candidate_a46ad7\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bcb796cd0bc8f0cdd0bcce9b:choice", "group_id": "fresh12:snli:bcb796cd0bc8f0cdd0bcce9b", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men stand near a piece of large machinery.\\nHypothesis: Two men are having sex\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e5bc8e\": \"The hypothesis follows from the premise.\", \"candidate_f71d16\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_7edbb0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e5bc8e\": 0.0, \"candidate_f71d16\": 0.0, \"candidate_7edbb0\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e4413793d9cff7a0f24d45ba:choice", "group_id": "fresh12:snli:e4413793d9cff7a0f24d45ba", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A group of people celebrating in the street.\\nHypothesis: People got shot by a sniper and dide\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ba2b9c\": \"The hypothesis follows from the premise.\", \"candidate_e3ce2c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d2684b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ba2b9c\": 0.0, \"candidate_e3ce2c\": 0.0, \"candidate_d2684b\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e211b5a2948ab1a67f1a4ee4:choice", "group_id": "fresh12:snli:e211b5a2948ab1a67f1a4ee4", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Businessman walking as two arrows point to him.\\nHypothesis: A guy is jogging in a sweatsuit.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_25d316\": \"The hypothesis follows from the premise.\", \"candidate_039f26\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8be819\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_25d316\": 0.0, \"candidate_039f26\": 0.0, \"candidate_8be819\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0092e9698cac98743318f1bc:choice", "group_id": "fresh12:snli:0092e9698cac98743318f1bc", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A barefoot child climbs at the jungle gym.\\nHypothesis: The child is sitting on a bed.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_53749d\": \"The hypothesis follows from the premise.\", \"candidate_69dcb8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_bac5df\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_53749d\": 0.0, \"candidate_69dcb8\": 0.0, \"candidate_bac5df\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d6082de7ba6a607113321928:choice", "group_id": "fresh12:snli:d6082de7ba6a607113321928", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Construction workers are standing upon on a wooden bridge.\\nHypothesis: The bridge is metal\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1a3720\": \"The hypothesis follows from the premise.\", \"candidate_851b5f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ed2f5c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1a3720\": 0.0, \"candidate_851b5f\": 0.0, \"candidate_ed2f5c\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6ee5b7a8eb493380f00012c5:choice", "group_id": "fresh12:snli:6ee5b7a8eb493380f00012c5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: construction worker working on a structure\\nHypothesis: A man is working on a computer.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7ecabe\": \"The hypothesis follows from the premise.\", \"candidate_7f5982\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1572b9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7ecabe\": 0.0, \"candidate_7f5982\": 0.0, \"candidate_1572b9\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9adcf58561d89542b1a68d84:choice", "group_id": "fresh12:snli:9adcf58561d89542b1a68d84", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A crowd of people in stadium style seating appear disinterested in the venue.\\nHypothesis: A group of people are excited to be here.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9ff579\": \"The hypothesis follows from the premise.\", \"candidate_930ad8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5da8f8\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9ff579\": 0.0, \"candidate_930ad8\": 0.0, \"candidate_5da8f8\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5b84234dbd41d3f9118aedbe:choice", "group_id": "fresh12:snli:5b84234dbd41d3f9118aedbe", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy in an all blue uniform kicks a blue soccer ball in a grassy area.\\nHypothesis: A boy is playing with a squirt gun.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_62d8c6\": \"The hypothesis follows from the premise.\", \"candidate_f8f226\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3cf95e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_62d8c6\": 0.0, \"candidate_f8f226\": 0.0, \"candidate_3cf95e\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9f8d088a595d88594bf3fadd:choice", "group_id": "fresh12:snli:9f8d088a595d88594bf3fadd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man stands by two face structures on Easter Island.\\nHypothesis: A child hunts for easter eggs.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_450d67\": \"The hypothesis follows from the premise.\", \"candidate_16a3bc\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2d5d9b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_450d67\": 0.0, \"candidate_16a3bc\": 0.0, \"candidate_2d5d9b\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:040b9905039857e08cf3f1a2:choice", "group_id": "fresh12:snli:040b9905039857e08cf3f1a2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The players manuever for the soccer ball.\\nHypothesis: Bowlers are eating sandwiches.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3bafd2\": \"The hypothesis follows from the premise.\", \"candidate_9098cf\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_bf6c28\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3bafd2\": 0.0, \"candidate_9098cf\": 0.0, \"candidate_bf6c28\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:dfdfa105df9a2150d7ee7710:choice", "group_id": "fresh12:snli:dfdfa105df9a2150d7ee7710", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Young boy wearing pajamas lying down in bed and sleeping on a colorful striped pillow.\\nHypothesis: A young boy is playing soccer.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3ae457\": \"The hypothesis follows from the premise.\", \"candidate_c9f2ab\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e2127e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3ae457\": 0.0, \"candidate_c9f2ab\": 0.0, \"candidate_e2127e\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c6842361834a390a631c91bd:choice", "group_id": "fresh12:snli:c6842361834a390a631c91bd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: an old shoemaker in his factory\\nHypothesis: The shoemaker is getting ready for his 16th birthday.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_edfb75\": \"The hypothesis follows from the premise.\", \"candidate_396b2c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5d68c9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_edfb75\": 0.0, \"candidate_396b2c\": 0.0, \"candidate_5d68c9\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:8105854650cf298facc47673:choice", "group_id": "fresh12:snli:8105854650cf298facc47673", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man running in a race passes by spectators.\\nHypothesis: A man in a kayak passes by spectators.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bf28d6\": \"The hypothesis follows from the premise.\", \"candidate_79c118\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2a2830\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bf28d6\": 0.0, \"candidate_79c118\": 0.0, \"candidate_2a2830\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b5db1606d6d241f96c08b93e:choice", "group_id": "fresh12:snli:b5db1606d6d241f96c08b93e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman doing gymnastics, she is using the balance beam\\nHypothesis: A woman is napping in the gym.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8d5ef2\": \"The hypothesis follows from the premise.\", \"candidate_a2ee3a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d12de2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8d5ef2\": 0.0, \"candidate_a2ee3a\": 0.0, \"candidate_d12de2\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:176e48f65eff671f7a7f5001:choice", "group_id": "fresh12:snli:176e48f65eff671f7a7f5001", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young male wearing a blue hat turned backwards and slightly sideways with a cigarette in his mouth.\\nHypothesis: A man wears his hat on his foot.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a2c47c\": \"The hypothesis follows from the premise.\", \"candidate_b0b648\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ac399a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a2c47c\": 0.0, \"candidate_b0b648\": 0.0, \"candidate_ac399a\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1281207b8c0ccd279cfc45b8:choice", "group_id": "fresh12:snli:1281207b8c0ccd279cfc45b8", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A lady is walking out of a short tunnel as a male prepares to enter it from the opposite side.\\nHypothesis: Everybody is sitting\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_df4a82\": \"The hypothesis follows from the premise.\", \"candidate_c76813\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_764971\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_df4a82\": 0.0, \"candidate_c76813\": 0.0, \"candidate_764971\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b864a400aaed2078c183d4b9:choice", "group_id": "fresh12:snli:b864a400aaed2078c183d4b9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A person in a ken-do outfit is standing there waiting.\\nHypothesis: Nobody is standing\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1087e7\": \"The hypothesis follows from the premise.\", \"candidate_cefefa\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_599080\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1087e7\": 0.0, \"candidate_cefefa\": 0.0, \"candidate_599080\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e503b8319898cb56c04acd5c:choice", "group_id": "fresh12:snli:e503b8319898cb56c04acd5c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy in red and a girl pink are walking through a low cut field.\\nHypothesis: Kids are walking through a mall.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cf799b\": \"The hypothesis follows from the premise.\", \"candidate_1e98c1\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d00341\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cf799b\": 0.0, \"candidate_1e98c1\": 0.0, \"candidate_d00341\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ead5d4d24515d71b717647e0:choice", "group_id": "fresh12:snli:ead5d4d24515d71b717647e0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A lady in a white religious gown is being questioned by many people as she walks.\\nHypothesis: nobody seen her when the angel came on earth\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6108c0\": \"The hypothesis follows from the premise.\", \"candidate_edbd5d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_428e7a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6108c0\": 0.0, \"candidate_edbd5d\": 0.0, \"candidate_428e7a\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d4ed620f44c207dd00e164bc:choice", "group_id": "fresh12:snli:d4ed620f44c207dd00e164bc", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a group of people dancing together.\\nHypothesis: No one is dancing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_11134c\": \"The hypothesis follows from the premise.\", \"candidate_0c799b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9777aa\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_11134c\": 0.0, \"candidate_0c799b\": 0.0, \"candidate_9777aa\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6654630c5185d3641d1ccd6c:choice", "group_id": "fresh12:snli:6654630c5185d3641d1ccd6c", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two men and one woman with blond-hair walk through a gate with barbed wire on top.\\nHypothesis: Two dogs are walking throught a gate.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2002e0\": \"The hypothesis follows from the premise.\", \"candidate_0ebe56\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_8340df\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2002e0\": 0.0, \"candidate_0ebe56\": 0.0, \"candidate_8340df\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:3b8ca5456ed5740d34d7d908:choice", "group_id": "fresh12:snli:3b8ca5456ed5740d34d7d908", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A grim looking man with sunglasses pilots a boat.\\nHypothesis: The happy pilot flies his plane.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f8dadf\": \"The hypothesis follows from the premise.\", \"candidate_f610a2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_80d13e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f8dadf\": 0.0, \"candidate_f610a2\": 0.0, \"candidate_80d13e\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:32e8597ebd2c9b4d8398383e:choice", "group_id": "fresh12:snli:32e8597ebd2c9b4d8398383e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A frowning old man in a military cap and a civilian suit stands amongst a crowd of people outdoors carrying signs and balloons.\\nHypothesis: A man is watching a group of people having a hot dog eating contest.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_df2a25\": \"The hypothesis follows from the premise.\", \"candidate_8f397b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_cdd0a6\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_df2a25\": 0.0, \"candidate_8f397b\": 0.0, \"candidate_cdd0a6\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cb5d7cf9a4a01e954fe88719:choice", "group_id": "fresh12:snli:cb5d7cf9a4a01e954fe88719", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A rescue worker searches with a flashlight under a slab in a ruined building.\\nHypothesis: A survivor searches for a meal.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_74b27a\": \"The hypothesis follows from the premise.\", \"candidate_e838db\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0fb7f9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_74b27a\": 0.0, \"candidate_e838db\": 0.0, \"candidate_0fb7f9\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:315492772e7b486c380d0259:choice", "group_id": "fresh12:snli:315492772e7b486c380d0259", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in purple tie is singing.\\nHypothesis: A man is talking to others.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7a4875\": \"The hypothesis follows from the premise.\", \"candidate_44e49c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_fbcb6b\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7a4875\": 0.0, \"candidate_44e49c\": 0.0, \"candidate_fbcb6b\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4ca344c15424a264cf3070ba:choice", "group_id": "fresh12:snli:4ca344c15424a264cf3070ba", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two workers in green uniforms are standing in an alleyway.\\nHypothesis: The workers are wearing black.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_65ad49\": \"The hypothesis follows from the premise.\", \"candidate_60ebd2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f0307f\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_65ad49\": 0.0, \"candidate_60ebd2\": 0.0, \"candidate_f0307f\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:5bd6b7e0b417371c06d38e49:choice", "group_id": "fresh12:snli:5bd6b7e0b417371c06d38e49", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a black hat opens his mouth.\\nHypothesis: A man in black hat keeps his lips sealed.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f7895c\": \"The hypothesis follows from the premise.\", \"candidate_c7a624\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9ab0ef\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f7895c\": 0.0, \"candidate_c7a624\": 0.0, \"candidate_9ab0ef\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bfd3ae1fb9ec29f6643acf6e:choice", "group_id": "fresh12:snli:bfd3ae1fb9ec29f6643acf6e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Man sitting playing guitar.\\nHypothesis: A man is standing on stage at a concert, playing saxophone.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a15793\": \"The hypothesis follows from the premise.\", \"candidate_1b457f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_05ab3c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a15793\": 0.0, \"candidate_1b457f\": 0.0, \"candidate_05ab3c\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e4f6b161687d1fb7d9529f38:choice", "group_id": "fresh12:snli:e4f6b161687d1fb7d9529f38", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Three men on mountain bikes go over a sand hill.\\nHypothesis: Men walk their bikes after getting flat tires.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dba7fa\": \"The hypothesis follows from the premise.\", \"candidate_ecce7e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0dd089\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dba7fa\": 0.0, \"candidate_ecce7e\": 0.0, \"candidate_0dd089\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:98f0f698813b2ed56051bd27:choice", "group_id": "fresh12:snli:98f0f698813b2ed56051bd27", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Small girl in an apron placing confetti in a woman's hair.\\nHypothesis: A woman bathing her daughter.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f94987\": \"The hypothesis follows from the premise.\", \"candidate_a2c8b5\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_17e504\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f94987\": 0.0, \"candidate_a2c8b5\": 0.0, \"candidate_17e504\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7deb55de062962ebc33d06d3:choice", "group_id": "fresh12:snli:7deb55de062962ebc33d06d3", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A brown and white dog swims in a backyard swimming pool.\\nHypothesis: A dog sleeping on the porch.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d3c8c7\": \"The hypothesis follows from the premise.\", \"candidate_7b742c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3d2d4c\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d3c8c7\": 0.0, \"candidate_7b742c\": 0.0, \"candidate_3d2d4c\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a910703f000f8d79c8ab22bb:choice", "group_id": "fresh12:snli:a910703f000f8d79c8ab22bb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Numerous customers browsing for produce in a market.\\nHypothesis: People are showering.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5656f0\": \"The hypothesis follows from the premise.\", \"candidate_410a0a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_41b527\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5656f0\": 0.0, \"candidate_410a0a\": 0.0, \"candidate_41b527\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:f4971de59dd34ad60498c5a1:choice", "group_id": "fresh12:snli:f4971de59dd34ad60498c5a1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman is chatting as she drinks her coffee.\\nHypothesis: The woman is silent.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f3f213\": \"The hypothesis follows from the premise.\", \"candidate_7a38ee\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4e4c40\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f3f213\": 0.0, \"candidate_7a38ee\": 0.0, \"candidate_4e4c40\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e334a71c92b374d870923696:choice", "group_id": "fresh12:snli:e334a71c92b374d870923696", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: We can see the white underwear of the man wearing the red soccer jersey because he is pulling up his shorts.\\nHypothesis: People are not looking at underwear.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8514f3\": \"The hypothesis follows from the premise.\", \"candidate_deed7b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_88f5c3\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8514f3\": 0.0, \"candidate_deed7b\": 0.0, \"candidate_88f5c3\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:fd3a6e344bca99aa7dad10e6:choice", "group_id": "fresh12:snli:fd3a6e344bca99aa7dad10e6", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man in a black shirt is looking at a bike in a workshop.\\nHypothesis: A man is wearing a red shirt\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6ba8da\": \"The hypothesis follows from the premise.\", \"candidate_e173b5\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_a81ebe\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6ba8da\": 0.0, \"candidate_e173b5\": 0.0, \"candidate_a81ebe\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:90ca5c9ce33e795d7724a0a1:choice", "group_id": "fresh12:snli:90ca5c9ce33e795d7724a0a1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man with a long white beard is examining a camera and another man with a black shirt is in the background.\\nHypothesis: A man takes a picture of a woman\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_86407f\": \"The hypothesis follows from the premise.\", \"candidate_0a5d35\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_4c9f45\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_86407f\": 0.0, \"candidate_0a5d35\": 0.0, \"candidate_4c9f45\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:3739a1ac7789c293de09604e:choice", "group_id": "fresh12:snli:3739a1ac7789c293de09604e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young boy runs across a road in front of a sky blue building with barred windows.\\nHypothesis: A boy sleeps in his bed.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bf29fc\": \"The hypothesis follows from the premise.\", \"candidate_8b5b55\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_42f597\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bf29fc\": 0.0, \"candidate_8b5b55\": 0.0, \"candidate_42f597\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:fefe0462a253a5018d5ab6fe:choice", "group_id": "fresh12:snli:fefe0462a253a5018d5ab6fe", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: The dog is in snow, and has a stick in its mouth.\\nHypothesis: The dog is sleeping on the couch.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f2f6a1\": \"The hypothesis follows from the premise.\", \"candidate_caa0c0\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_11b7d5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f2f6a1\": 0.0, \"candidate_caa0c0\": 0.0, \"candidate_11b7d5\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:116d35b79378ff8782da4892:choice", "group_id": "fresh12:snli:116d35b79378ff8782da4892", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A football layer wearing a red shirt.\\nHypothesis: A football player in a black jersey.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6f9c32\": \"The hypothesis follows from the premise.\", \"candidate_8afa6b\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0ee248\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6f9c32\": 0.0, \"candidate_8afa6b\": 0.0, \"candidate_0ee248\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:52865f75a34947b11452da1e:choice", "group_id": "fresh12:snli:52865f75a34947b11452da1e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A boy scout with a red cap is looking through a telescope\\nHypothesis: A boy scout is making a campfire.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0ab78d\": \"The hypothesis follows from the premise.\", \"candidate_5d61e5\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_55f46d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0ab78d\": 0.0, \"candidate_5d61e5\": 0.0, \"candidate_55f46d\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cd18842de1f68a4703aa688a:choice", "group_id": "fresh12:snli:cd18842de1f68a4703aa688a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: It looks like the clown has fallen off the horse.\\nHypothesis: The cowboy rode the bull to the buzzer.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5d33c7\": \"The hypothesis follows from the premise.\", \"candidate_f89471\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0151da\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5d33c7\": 0.0, \"candidate_f89471\": 0.0, \"candidate_0151da\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:37891adba92196214835c0ff:choice", "group_id": "fresh12:snli:37891adba92196214835c0ff", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A U.S. soldier helps a young girl in the hands of her father, while other soldiers look on.\\nHypothesis: The girl is by herself.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a19a23\": \"The hypothesis follows from the premise.\", \"candidate_2ecc27\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0bcd0a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a19a23\": 0.0, \"candidate_2ecc27\": 0.0, \"candidate_0bcd0a\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:d8fbe1d866fef0a978ff2b36:choice", "group_id": "fresh12:snli:d8fbe1d866fef0a978ff2b36", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two sweet young girls on ice skates, hands up and holding on to each other get ready to begin their performance on the ice.\\nHypothesis: Two female synchronized swimmers are in the pool.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d70dcd\": \"The hypothesis follows from the premise.\", \"candidate_3eeec8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f4b894\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d70dcd\": 0.0, \"candidate_3eeec8\": 0.0, \"candidate_f4b894\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6dbd151e9ca6ab7bb2ca398e:choice", "group_id": "fresh12:snli:6dbd151e9ca6ab7bb2ca398e", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A young woman in a black skirt and boots walks by stores.\\nHypothesis: A young woman is eating in a fine restaurant.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f590c4\": \"The hypothesis follows from the premise.\", \"candidate_30e031\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2d3af2\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f590c4\": 0.0, \"candidate_30e031\": 0.0, \"candidate_2d3af2\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:59c44860b07720974fcd58cb:choice", "group_id": "fresh12:snli:59c44860b07720974fcd58cb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A homeless man carries a sign that says \\\"hungry\\\".\\nHypothesis: The man is living in a mansion.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0259c1\": \"The hypothesis follows from the premise.\", \"candidate_870577\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_3fa808\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0259c1\": 0.0, \"candidate_870577\": 0.0, \"candidate_3fa808\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:e327519857acf98937d5abf2:choice", "group_id": "fresh12:snli:e327519857acf98937d5abf2", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man holding girl at top of a slide\\nHypothesis: A man with a girl on a swing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_00f5ee\": \"The hypothesis follows from the premise.\", \"candidate_6efb1c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f789e5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_00f5ee\": 0.0, \"candidate_6efb1c\": 0.0, \"candidate_f789e5\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:4be16176b1644c7f1f81689f:choice", "group_id": "fresh12:snli:4be16176b1644c7f1f81689f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A member of a Bergwacht team leads a sled laden with orange material while on skis as his colleague stands in the foreground.\\nHypothesis: A man is snowboarding.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2ff871\": \"The hypothesis follows from the premise.\", \"candidate_0ad323\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_beabcd\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2ff871\": 0.0, \"candidate_0ad323\": 0.0, \"candidate_beabcd\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c14b98af0ce5a7c0e27fda12:choice", "group_id": "fresh12:snli:c14b98af0ce5a7c0e27fda12", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Cheerleaders in blue performing on a football field underneath a yellow football goal post.\\nHypothesis: The San Francisco 49ers cheerleaders perform in a dark basement.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_67ecf3\": \"The hypothesis follows from the premise.\", \"candidate_b307d8\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c54c25\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_67ecf3\": 0.0, \"candidate_b307d8\": 0.0, \"candidate_c54c25\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6d9bd26164ab4edfdc937589:choice", "group_id": "fresh12:snli:6d9bd26164ab4edfdc937589", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: People seated at long tables all facing the same direction, some writing and some watching.\\nHypothesis: A group of people are sitting at the tables sharing a meal.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_63d6d3\": \"The hypothesis follows from the premise.\", \"candidate_79c979\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_1bee48\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_63d6d3\": 0.0, \"candidate_79c979\": 0.0, \"candidate_1bee48\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:691aaab413426caee0e9f754:choice", "group_id": "fresh12:snli:691aaab413426caee0e9f754", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A guy with a football, a crowd watching him.\\nHypothesis: There is a guy with a basketball.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5448da\": \"The hypothesis follows from the premise.\", \"candidate_cd9970\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_445556\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5448da\": 0.0, \"candidate_cd9970\": 0.0, \"candidate_445556\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:cd91f9392fae4d5de40ca398:choice", "group_id": "fresh12:snli:cd91f9392fae4d5de40ca398", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Men sitting at a table in a church.\\nHypothesis: The men pray at the alters.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4758fa\": \"The hypothesis follows from the premise.\", \"candidate_b9f2bc\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_93bc3a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4758fa\": 0.0, \"candidate_b9f2bc\": 0.0, \"candidate_93bc3a\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c090c2a59b862e961ae4d6fd:choice", "group_id": "fresh12:snli:c090c2a59b862e961ae4d6fd", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a man wearing a red jacket and camouflage pants is leaning against a giant tree in the snow.\\nHypothesis: A man sips beer on the beach\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_98d94f\": \"The hypothesis follows from the premise.\", \"candidate_bdebf5\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_473bb3\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_98d94f\": 0.0, \"candidate_bdebf5\": 0.0, \"candidate_473bb3\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:bf52af1fd2f1c868f038f0ff:choice", "group_id": "fresh12:snli:bf52af1fd2f1c868f038f0ff", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two boys run to the entryway of an old building.\\nHypothesis: A couple of children are eating popsicles at a birthday party.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2b0ecc\": \"The hypothesis follows from the premise.\", \"candidate_cad5fb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_48d332\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2b0ecc\": 0.0, \"candidate_cad5fb\": 0.0, \"candidate_48d332\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:618053e2277b795002e7ef60:choice", "group_id": "fresh12:snli:618053e2277b795002e7ef60", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Smilling and laughing baby in a walker with fire engine airplane mobile and other toys on the floor\\nHypothesis: The baby's diaper is full of feces.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_04eb5d\": \"The hypothesis follows from the premise.\", \"candidate_780eee\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_2c63e5\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_04eb5d\": 0.0, \"candidate_780eee\": 0.0, \"candidate_2c63e5\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:0844da5c07a9ed9c11020fd9:choice", "group_id": "fresh12:snli:0844da5c07a9ed9c11020fd9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Two women on a beach.\\nHypothesis: Two women are resting indoors.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_70f8a2\": \"The hypothesis follows from the premise.\", \"candidate_dac923\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_0f7c86\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_70f8a2\": 0.0, \"candidate_dac923\": 0.0, \"candidate_0f7c86\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:9a555b684d0202a36ce84266:choice", "group_id": "fresh12:snli:9a555b684d0202a36ce84266", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: a woman walking down the street.\\nHypothesis: A woman rides a bike.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eaa62c\": \"The hypothesis follows from the premise.\", \"candidate_4e4d3c\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_b506fb\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eaa62c\": 0.0, \"candidate_4e4d3c\": 0.0, \"candidate_b506fb\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:659d8e40c2e368d07d0d4ca0:choice", "group_id": "fresh12:snli:659d8e40c2e368d07d0d4ca0", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: GEntleman sketching on brown paper at Christmas.\\nHypothesis: The woman drew on canvas.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f6d6ea\": \"The hypothesis follows from the premise.\", \"candidate_880d6f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e08807\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f6d6ea\": 0.0, \"candidate_880d6f\": 0.0, \"candidate_e08807\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:1cde5bd71f38188f166d34bb:choice", "group_id": "fresh12:snli:1cde5bd71f38188f166d34bb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man and a woman, both wearing sunglasses, embrace one another, while smiling and posing for the camera.\\nHypothesis: Two men take a picture together.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3d6e30\": \"The hypothesis follows from the premise.\", \"candidate_81194d\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_69d399\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3d6e30\": 0.0, \"candidate_81194d\": 0.0, \"candidate_69d399\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:7dc5991e1949aa1fb13c716f:choice", "group_id": "fresh12:snli:7dc5991e1949aa1fb13c716f", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Spectators look on as a baseball player runs onto a base while the baseman waits to make the catch.\\nHypothesis: a cat jumps high\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c02f6f\": \"The hypothesis follows from the premise.\", \"candidate_207d0f\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_90f8da\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c02f6f\": 0.0, \"candidate_207d0f\": 0.0, \"candidate_90f8da\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b401d85524a76655bb1464d1:choice", "group_id": "fresh12:snli:b401d85524a76655bb1464d1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Mother and daughter are meeting for the first time.\\nHypothesis: The mother and daughter are fighitn.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_56ee73\": \"The hypothesis follows from the premise.\", \"candidate_d7cef4\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_20d0ff\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_56ee73\": 0.0, \"candidate_d7cef4\": 0.0, \"candidate_20d0ff\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:c5936a10092f7e98f6531f86:choice", "group_id": "fresh12:snli:c5936a10092f7e98f6531f86", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man wearing a hat and boots is digging for something in the snow.\\nHypothesis: A man is walking on sunshine.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a75a80\": \"The hypothesis follows from the premise.\", \"candidate_e809fb\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_5fa655\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a75a80\": 0.0, \"candidate_e809fb\": 0.0, \"candidate_5fa655\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:164da7a02f4d30fa0abcc125:choice", "group_id": "fresh12:snli:164da7a02f4d30fa0abcc125", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man wearing red sunglasses is carrying two cameras.\\nHypothesis: The man has blue sunglasses.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_190038\": \"The hypothesis follows from the premise.\", \"candidate_a6f205\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_27fe73\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_190038\": 0.0, \"candidate_a6f205\": 0.0, \"candidate_27fe73\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:54bb970fc91b522651d58548:choice", "group_id": "fresh12:snli:54bb970fc91b522651d58548", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A person is posed with one leg raised and one arm out at a competition with their teammates in the background.\\nHypothesis: A man posed with both arms out at a competition.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_12af9e\": \"The hypothesis follows from the premise.\", \"candidate_e27570\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_48fc12\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_12af9e\": 0.0, \"candidate_e27570\": 0.0, \"candidate_48fc12\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:73cf2b2396ba0898e4ab5520:choice", "group_id": "fresh12:snli:73cf2b2396ba0898e4ab5520", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A view of bicyclists on a city street, riding in both directions, with red reflective safety cones dividing the two lanes.\\nHypothesis: Bumper to bumper automobile traffic clogs the street.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_04b5b7\": \"The hypothesis follows from the premise.\", \"candidate_9c929e\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f34bac\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_04b5b7\": 0.0, \"candidate_9c929e\": 0.0, \"candidate_f34bac\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:eaf707b5534cdd22c3a0a225:choice", "group_id": "fresh12:snli:eaf707b5534cdd22c3a0a225", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A large group of people walking in a busy city at night.\\nHypothesis: People are eating lunch in a restaurant.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_785bf5\": \"The hypothesis follows from the premise.\", \"candidate_4ce011\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_c9efc9\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_785bf5\": 0.0, \"candidate_4ce011\": 0.0, \"candidate_c9efc9\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:b2b412c1164b4dbe759f07bb:choice", "group_id": "fresh12:snli:b2b412c1164b4dbe759f07bb", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A fat woman wearing faint blue top is blowing soap bubble.\\nHypothesis: A fat woman is topless.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_28ec97\": \"The hypothesis follows from the premise.\", \"candidate_8a2e96\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_9a224d\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_28ec97\": 0.0, \"candidate_8a2e96\": 0.0, \"candidate_9a224d\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ef4637806d899876d4f706b5:choice", "group_id": "fresh12:snli:ef4637806d899876d4f706b5", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An ATV rides over a vast sandy landscape.\\nHypothesis: The rider is in his garage\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b0cfb5\": \"The hypothesis follows from the premise.\", \"candidate_824df7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_f56f64\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b0cfb5\": 0.0, \"candidate_824df7\": 0.0, \"candidate_f56f64\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:3ab62c0eacda8871724ad256:choice", "group_id": "fresh12:snli:3ab62c0eacda8871724ad256", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Women are sitting around a table at a restaurant chatting.\\nHypothesis: Women are knitting scarves for the homeless.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_55afaf\": \"The hypothesis follows from the premise.\", \"candidate_831915\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_29b533\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_55afaf\": 0.0, \"candidate_831915\": 0.0, \"candidate_29b533\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:6a3a48e8830ce9af12ad062a:choice", "group_id": "fresh12:snli:6a3a48e8830ce9af12ad062a", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: An asian teenager wearing a white shirt with black stripes is sitting on the steps with his face resting on his hand.\\nHypothesis: The teenager is sliding down the railing.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_707bb4\": \"The hypothesis follows from the premise.\", \"candidate_53bdde\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_da066a\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_707bb4\": 0.0, \"candidate_53bdde\": 0.0, \"candidate_da066a\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ee9b7a3cd62986eaccd0d8a9:choice", "group_id": "fresh12:snli:ee9b7a3cd62986eaccd0d8a9", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man falling off a bull as the animal jumps into the air.\\nHypothesis: A man sits on the calm bull.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d17213\": \"The hypothesis follows from the premise.\", \"candidate_86d010\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_914862\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d17213\": 0.0, \"candidate_86d010\": 0.0, \"candidate_914862\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:a177cca6145d977c8390becf:choice", "group_id": "fresh12:snli:a177cca6145d977c8390becf", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in a teal apron prepares a meal at a restaurant.\\nHypothesis: A woman is walking in park\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ebb85e\": \"The hypothesis follows from the premise.\", \"candidate_827ce7\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_d2fd02\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ebb85e\": 0.0, \"candidate_827ce7\": 0.0, \"candidate_d2fd02\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:424f8e9c601399b21f956a25:choice", "group_id": "fresh12:snli:424f8e9c601399b21f956a25", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A woman in a blue uniform, with long curly red-hair, looking at her horse before the race.\\nHypothesis: The horse is dead.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1c5e75\": \"The hypothesis follows from the premise.\", \"candidate_36cd46\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_26afc0\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1c5e75\": 0.0, \"candidate_36cd46\": 0.0, \"candidate_26afc0\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:63ef89c40ce1cd27afd9b7e1:choice", "group_id": "fresh12:snli:63ef89c40ce1cd27afd9b7e1", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: A man with a blue helmet and orange shirt.\\nHypothesis: A woman has a helmet on a motorcycle.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3de136\": \"The hypothesis follows from the premise.\", \"candidate_9ddb8a\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_e9e326\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3de136\": 0.0, \"candidate_9ddb8a\": 0.0, \"candidate_e9e326\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:snli:ada0ecef293d6ba293249b05:choice", "group_id": "fresh12:snli:ada0ecef293d6ba293249b05", "split": "test", "family": "text_nli", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Premise: Firemen stand outside of their firetruck with all its lights on.\\nHypothesis: Policemen stand outside of their firetruck with all its lights on.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c1231c\": \"The hypothesis follows from the premise.\", \"candidate_b073f2\": \"The premise does not determine whether the hypothesis is true.\", \"candidate_ecbd2e\": \"The hypothesis contradicts the premise.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c1231c\": 0.0, \"candidate_b073f2\": 0.0, \"candidate_ecbd2e\": 1.0}}", "source_id": "stanfordnlp/snli", "source_revision": "cdb5c3d5eed6ead6e5a341c8e56e669bb666725b", "source_license": "CC-BY-SA-4.0", "source_url": "https://huggingface.co/datasets/stanfordnlp/snli", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8fe8acafcca5748bcb439676:choice", "group_id": "fresh12:banking77:8fe8acafcca5748bcb439676", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I'm still waiting on my card to be delivered.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_198ce7\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_5f8aaf\": \"The user's request concerns card payment not recognised.\", \"candidate_de60de\": \"The user's request concerns card arrival.\", \"candidate_9a0d58\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_23c354\": \"The user's request concerns getting virtual card.\", \"candidate_6f4052\": \"The user's request concerns terminate account.\", \"candidate_860b97\": \"The user's request concerns card acceptance.\", \"candidate_80cb7c\": \"The user's request concerns transfer fee charged.\"}}}}", "targets_json": "{\"decision\": {\"candidate_198ce7\": 0.0, \"candidate_5f8aaf\": 0.0, \"candidate_de60de\": 1.0, \"candidate_9a0d58\": 0.0, \"candidate_23c354\": 0.0, \"candidate_6f4052\": 0.0, \"candidate_860b97\": 0.0, \"candidate_80cb7c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:bef33f3f3d6d70208cc533e5:choice", "group_id": "fresh12:banking77:bef33f3f3d6d70208cc533e5", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card hasn't arrived yet.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b9e07e\": \"The user's request concerns country support.\", \"candidate_32a1a5\": \"The user's request concerns virtual card not working.\", \"candidate_27bd37\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_ee3584\": \"The user's request concerns cancel transfer.\", \"candidate_8b7818\": \"The user's request concerns card arrival.\", \"candidate_8b3b2b\": \"The user's request concerns beneficiary not allowed.\", \"candidate_b24bf0\": \"The user's request concerns card swallowed.\", \"candidate_7bb441\": \"The user's request concerns top up by card charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b9e07e\": 0.0, \"candidate_32a1a5\": 0.0, \"candidate_27bd37\": 0.0, \"candidate_ee3584\": 0.0, \"candidate_8b7818\": 1.0, \"candidate_8b3b2b\": 0.0, \"candidate_b24bf0\": 0.0, \"candidate_7bb441\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b0f48e018de594d46e7a1ff4:choice", "group_id": "fresh12:banking77:b0f48e018de594d46e7a1ff4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I think something went wrong with my card delivery as I haven't received it yet.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b341a0\": \"The user's request concerns card acceptance.\", \"candidate_d5f0b6\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_f5a9fe\": \"The user's request concerns topping up by card.\", \"candidate_9d686e\": \"The user's request concerns top up by card charge.\", \"candidate_f1b78f\": \"The user's request concerns compromised card.\", \"candidate_72562c\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_125dc5\": \"The user's request concerns reverted card payment?.\", \"candidate_2c7c00\": \"The user's request concerns card arrival.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b341a0\": 0.0, \"candidate_d5f0b6\": 0.0, \"candidate_f5a9fe\": 0.0, \"candidate_9d686e\": 0.0, \"candidate_f1b78f\": 0.0, \"candidate_72562c\": 0.0, \"candidate_125dc5\": 0.0, \"candidate_2c7c00\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ef454b878a137bb465e1f96d:choice", "group_id": "fresh12:banking77:ef454b878a137bb465e1f96d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What is the tracking number for the card you sent?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_304dbe\": \"The user's request concerns verify my identity.\", \"candidate_157226\": \"The user's request concerns request refund.\", \"candidate_625535\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_3cb664\": \"The user's request concerns pending top up.\", \"candidate_422e9a\": \"The user's request concerns card arrival.\", \"candidate_a86af4\": \"The user's request concerns passcode forgotten.\", \"candidate_d7cae8\": \"The user's request concerns terminate account.\", \"candidate_14524d\": \"The user's request concerns declined transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_304dbe\": 0.0, \"candidate_157226\": 0.0, \"candidate_625535\": 0.0, \"candidate_3cb664\": 0.0, \"candidate_422e9a\": 1.0, \"candidate_a86af4\": 0.0, \"candidate_d7cae8\": 0.0, \"candidate_14524d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4c5bdb732c12ffdf9405e904:choice", "group_id": "fresh12:banking77:4c5bdb732c12ffdf9405e904", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I ordered a card but it has not arrived. Help please!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ff2acd\": \"The user's request concerns card payment fee charged.\", \"candidate_692d1e\": \"The user's request concerns card delivery estimate.\", \"candidate_118111\": \"The user's request concerns transfer not received by recipient.\", \"candidate_205397\": \"The user's request concerns disposable card limits.\", \"candidate_b85f8f\": \"The user's request concerns fiat currency support.\", \"candidate_97864f\": \"The user's request concerns contactless not working.\", \"candidate_5f613d\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_5c8183\": \"The user's request concerns card arrival.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ff2acd\": 0.0, \"candidate_692d1e\": 0.0, \"candidate_118111\": 0.0, \"candidate_205397\": 0.0, \"candidate_b85f8f\": 0.0, \"candidate_97864f\": 0.0, \"candidate_5f613d\": 0.0, \"candidate_5c8183\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:29532c035e96f93ebe438de5:choice", "group_id": "fresh12:banking77:29532c035e96f93ebe438de5", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I ordered a card a week ago, and it's still not here. What do I do?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_edd5af\": \"The user's request concerns failed transfer.\", \"candidate_6dc60e\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_22603d\": \"The user's request concerns declined card payment.\", \"candidate_c20456\": \"The user's request concerns contactless not working.\", \"candidate_2b8410\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_aa0cd0\": \"The user's request concerns activate my card.\", \"candidate_41dd30\": \"The user's request concerns card arrival.\", \"candidate_67addf\": \"The user's request concerns pending top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_edd5af\": 0.0, \"candidate_6dc60e\": 0.0, \"candidate_22603d\": 0.0, \"candidate_c20456\": 0.0, \"candidate_2b8410\": 0.0, \"candidate_aa0cd0\": 0.0, \"candidate_41dd30\": 1.0, \"candidate_67addf\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:087cfc9dd74c88142a8ddd5c:choice", "group_id": "fresh12:banking77:087cfc9dd74c88142a8ddd5c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I would like to reactivate my card.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7e5fef\": \"The user's request concerns atm support.\", \"candidate_7a1da8\": \"The user's request concerns request refund.\", \"candidate_90e92c\": \"The user's request concerns card payment fee charged.\", \"candidate_7fdd6d\": \"The user's request concerns pin blocked.\", \"candidate_470033\": \"The user's request concerns card linking.\", \"candidate_e2f7d8\": \"The user's request concerns declined card payment.\", \"candidate_a1df57\": \"The user's request concerns verify top up.\", \"candidate_c158b0\": \"The user's request concerns top up limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7e5fef\": 0.0, \"candidate_7a1da8\": 0.0, \"candidate_90e92c\": 0.0, \"candidate_7fdd6d\": 0.0, \"candidate_470033\": 1.0, \"candidate_e2f7d8\": 0.0, \"candidate_a1df57\": 0.0, \"candidate_c158b0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ad65a4115d52d21a36e1d146:choice", "group_id": "fresh12:banking77:ad65a4115d52d21a36e1d146", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Could you help me reactivate my card? It was previously lost, but I found it this morning in my jacket.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_28ca7b\": \"The user's request concerns getting spare card.\", \"candidate_4470be\": \"The user's request concerns card linking.\", \"candidate_82cdbb\": \"The user's request concerns transfer not received by recipient.\", \"candidate_c9ecc4\": \"The user's request concerns card acceptance.\", \"candidate_578dba\": \"The user's request concerns card payment fee charged.\", \"candidate_496c1c\": \"The user's request concerns card arrival.\", \"candidate_8c7401\": \"The user's request concerns compromised card.\", \"candidate_81917d\": \"The user's request concerns visa or mastercard.\"}}}}", "targets_json": "{\"decision\": {\"candidate_28ca7b\": 0.0, \"candidate_4470be\": 1.0, \"candidate_82cdbb\": 0.0, \"candidate_c9ecc4\": 0.0, \"candidate_578dba\": 0.0, \"candidate_496c1c\": 0.0, \"candidate_8c7401\": 0.0, \"candidate_81917d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ca4ad896fadedf6350891839:choice", "group_id": "fresh12:banking77:ca4ad896fadedf6350891839", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I reactivate a card I thought I lost?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b1402e\": \"The user's request concerns fiat currency support.\", \"candidate_94dc6b\": \"The user's request concerns lost or stolen phone.\", \"candidate_1dfdaf\": \"The user's request concerns transfer fee charged.\", \"candidate_c8a2f5\": \"The user's request concerns verify top up.\", \"candidate_e622a3\": \"The user's request concerns country support.\", \"candidate_abd323\": \"The user's request concerns card linking.\", \"candidate_8a3355\": \"The user's request concerns exchange via app.\", \"candidate_0237eb\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b1402e\": 0.0, \"candidate_94dc6b\": 0.0, \"candidate_1dfdaf\": 0.0, \"candidate_c8a2f5\": 0.0, \"candidate_e622a3\": 0.0, \"candidate_abd323\": 1.0, \"candidate_8a3355\": 0.0, \"candidate_0237eb\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f9eb1a469060ad1d6196accd:choice", "group_id": "fresh12:banking77:f9eb1a469060ad1d6196accd", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I found my card, I would like to reactivate it.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_52a5fe\": \"The user's request concerns compromised card.\", \"candidate_740281\": \"The user's request concerns card linking.\", \"candidate_e57cce\": \"The user's request concerns transfer not received by recipient.\", \"candidate_3fac09\": \"The user's request concerns lost or stolen card.\", \"candidate_26aa8c\": \"The user's request concerns passcode forgotten.\", \"candidate_b68955\": \"The user's request concerns edit personal details.\", \"candidate_83b5f0\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_65f12d\": \"The user's request concerns automatic top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_52a5fe\": 0.0, \"candidate_740281\": 1.0, \"candidate_e57cce\": 0.0, \"candidate_3fac09\": 0.0, \"candidate_26aa8c\": 0.0, \"candidate_b68955\": 0.0, \"candidate_83b5f0\": 0.0, \"candidate_65f12d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7a14871930b4befbf178180a:choice", "group_id": "fresh12:banking77:7a14871930b4befbf178180a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I link a new card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f5f57c\": \"The user's request concerns receiving money.\", \"candidate_5af12c\": \"The user's request concerns transfer not received by recipient.\", \"candidate_88c885\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_7825cc\": \"The user's request concerns apple pay or google pay.\", \"candidate_3e9435\": \"The user's request concerns fiat currency support.\", \"candidate_69c757\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_964fea\": \"The user's request concerns card linking.\", \"candidate_05f2d8\": \"The user's request concerns pending transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f5f57c\": 0.0, \"candidate_5af12c\": 0.0, \"candidate_88c885\": 0.0, \"candidate_7825cc\": 0.0, \"candidate_3e9435\": 0.0, \"candidate_69c757\": 0.0, \"candidate_964fea\": 1.0, \"candidate_05f2d8\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a85de2d7e00b5efd100d4007:choice", "group_id": "fresh12:banking77:a85de2d7e00b5efd100d4007", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I link to my credit card with you?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_438699\": \"The user's request concerns card linking.\", \"candidate_7169de\": \"The user's request concerns card not working.\", \"candidate_8490ad\": \"The user's request concerns declined cash withdrawal.\", \"candidate_2776e7\": \"The user's request concerns transfer not received by recipient.\", \"candidate_5f21bd\": \"The user's request concerns card about to expire.\", \"candidate_90111c\": \"The user's request concerns wrong amount of cash received.\", \"candidate_b840fc\": \"The user's request concerns pending transfer.\", \"candidate_b37297\": \"The user's request concerns balance not updated after cheque or cash deposit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_438699\": 1.0, \"candidate_7169de\": 0.0, \"candidate_8490ad\": 0.0, \"candidate_2776e7\": 0.0, \"candidate_5f21bd\": 0.0, \"candidate_90111c\": 0.0, \"candidate_b840fc\": 0.0, \"candidate_b37297\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:70305f6fa66330596908e4fb:choice", "group_id": "fresh12:banking77:70305f6fa66330596908e4fb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where do you get your exchange rates from?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_70f017\": \"The user's request concerns age limit.\", \"candidate_75ecdf\": \"The user's request concerns verify my identity.\", \"candidate_af17c8\": \"The user's request concerns why verify identity.\", \"candidate_0ef193\": \"The user's request concerns exchange rate.\", \"candidate_92e0c1\": \"The user's request concerns Refund not showing up.\", \"candidate_39af7e\": \"The user's request concerns terminate account.\", \"candidate_2be474\": \"The user's request concerns transfer into account.\", \"candidate_dc182e\": \"The user's request concerns get physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_70f017\": 0.0, \"candidate_75ecdf\": 0.0, \"candidate_af17c8\": 0.0, \"candidate_0ef193\": 1.0, \"candidate_92e0c1\": 0.0, \"candidate_39af7e\": 0.0, \"candidate_2be474\": 0.0, \"candidate_dc182e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6b1e22d9a482361b1098e2b1:choice", "group_id": "fresh12:banking77:6b1e22d9a482361b1098e2b1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"From where are you getting your exchange rates?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c56fbe\": \"The user's request concerns top up by cash or cheque.\", \"candidate_c401b3\": \"The user's request concerns verify my identity.\", \"candidate_bf929e\": \"The user's request concerns exchange rate.\", \"candidate_a62970\": \"The user's request concerns country support.\", \"candidate_0e979b\": \"The user's request concerns why verify identity.\", \"candidate_ea68db\": \"The user's request concerns pending card payment.\", \"candidate_38349c\": \"The user's request concerns order physical card.\", \"candidate_4e65d8\": \"The user's request concerns transaction charged twice.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c56fbe\": 0.0, \"candidate_c401b3\": 0.0, \"candidate_bf929e\": 1.0, \"candidate_a62970\": 0.0, \"candidate_0e979b\": 0.0, \"candidate_ea68db\": 0.0, \"candidate_38349c\": 0.0, \"candidate_4e65d8\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:19840db7d586ef59f5ebe8b6:choice", "group_id": "fresh12:banking77:19840db7d586ef59f5ebe8b6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is it a good time to exchange?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b3ea57\": \"The user's request concerns lost or stolen phone.\", \"candidate_18b032\": \"The user's request concerns top up limits.\", \"candidate_a17065\": \"The user's request concerns transaction charged twice.\", \"candidate_b32fb8\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_d4210e\": \"The user's request concerns reverted card payment?.\", \"candidate_49be39\": \"The user's request concerns exchange rate.\", \"candidate_8a4c9c\": \"The user's request concerns pin blocked.\", \"candidate_6dd335\": \"The user's request concerns terminate account.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b3ea57\": 0.0, \"candidate_18b032\": 0.0, \"candidate_a17065\": 0.0, \"candidate_b32fb8\": 0.0, \"candidate_d4210e\": 0.0, \"candidate_49be39\": 1.0, \"candidate_8a4c9c\": 0.0, \"candidate_6dd335\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:51aad847027db706782afba1:choice", "group_id": "fresh12:banking77:51aad847027db706782afba1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What kind of foreign exchange rate will I get when I exchange my money?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c18a49\": \"The user's request concerns disposable card limits.\", \"candidate_1ff983\": \"The user's request concerns exchange rate.\", \"candidate_990572\": \"The user's request concerns beneficiary not allowed.\", \"candidate_8167a4\": \"The user's request concerns declined card payment.\", \"candidate_87bd3f\": \"The user's request concerns pending top up.\", \"candidate_e7172d\": \"The user's request concerns declined transfer.\", \"candidate_e7616a\": \"The user's request concerns contactless not working.\", \"candidate_93b33b\": \"The user's request concerns cash withdrawal charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c18a49\": 0.0, \"candidate_1ff983\": 1.0, \"candidate_990572\": 0.0, \"candidate_8167a4\": 0.0, \"candidate_87bd3f\": 0.0, \"candidate_e7172d\": 0.0, \"candidate_e7616a\": 0.0, \"candidate_93b33b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5f3fea431be698f91c1dc8d8:choice", "group_id": "fresh12:banking77:5f3fea431be698f91c1dc8d8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The exchange rate would be?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4728e2\": \"The user's request concerns country support.\", \"candidate_1bcb93\": \"The user's request concerns exchange rate.\", \"candidate_19dcf6\": \"The user's request concerns get physical card.\", \"candidate_143e20\": \"The user's request concerns disposable card limits.\", \"candidate_f250f0\": \"The user's request concerns age limit.\", \"candidate_824e31\": \"The user's request concerns unable to verify identity.\", \"candidate_89d025\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_6fbf6d\": \"The user's request concerns request refund.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4728e2\": 0.0, \"candidate_1bcb93\": 1.0, \"candidate_19dcf6\": 0.0, \"candidate_143e20\": 0.0, \"candidate_f250f0\": 0.0, \"candidate_824e31\": 0.0, \"candidate_89d025\": 0.0, \"candidate_6fbf6d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:94546c2b949f5cd4f08c756f:choice", "group_id": "fresh12:banking77:94546c2b949f5cd4f08c756f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can you explain your exchange rate policy to me?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d8bfb7\": \"The user's request concerns order physical card.\", \"candidate_cf02ba\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_c590f5\": \"The user's request concerns transfer not received by recipient.\", \"candidate_c356f1\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_8deae2\": \"The user's request concerns exchange rate.\", \"candidate_97edd9\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_d281f7\": \"The user's request concerns card delivery estimate.\", \"candidate_42279d\": \"The user's request concerns exchange charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d8bfb7\": 0.0, \"candidate_cf02ba\": 0.0, \"candidate_c590f5\": 0.0, \"candidate_c356f1\": 0.0, \"candidate_8deae2\": 1.0, \"candidate_97edd9\": 0.0, \"candidate_d281f7\": 0.0, \"candidate_42279d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8a5dd8e85c6e852b2caf7352:choice", "group_id": "fresh12:banking77:8a5dd8e85c6e852b2caf7352", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The exchange rate seems off on this transaction\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d1bc97\": \"The user's request concerns card linking.\", \"candidate_90c1c3\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_8f9941\": \"The user's request concerns transaction charged twice.\", \"candidate_e3fa3f\": \"The user's request concerns country support.\", \"candidate_59391c\": \"The user's request concerns topping up by card.\", \"candidate_5583a8\": \"The user's request concerns edit personal details.\", \"candidate_ab9c5e\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_ece8f1\": \"The user's request concerns card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d1bc97\": 0.0, \"candidate_90c1c3\": 1.0, \"candidate_8f9941\": 0.0, \"candidate_e3fa3f\": 0.0, \"candidate_59391c\": 0.0, \"candidate_5583a8\": 0.0, \"candidate_ab9c5e\": 0.0, \"candidate_ece8f1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8ece7392409546ccd43c8844:choice", "group_id": "fresh12:banking77:8ece7392409546ccd43c8844", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I bought something, but the exchange rate looks wrong.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eff817\": \"The user's request concerns failed transfer.\", \"candidate_b0af54\": \"The user's request concerns card swallowed.\", \"candidate_f53010\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_5f5620\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_5eeb5c\": \"The user's request concerns receiving money.\", \"candidate_85afc1\": \"The user's request concerns declined card payment.\", \"candidate_206169\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_457794\": \"The user's request concerns extra charge on statement.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eff817\": 0.0, \"candidate_b0af54\": 0.0, \"candidate_f53010\": 0.0, \"candidate_5f5620\": 0.0, \"candidate_5eeb5c\": 0.0, \"candidate_85afc1\": 0.0, \"candidate_206169\": 1.0, \"candidate_457794\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ccad56657a290caa7bbe52f6:choice", "group_id": "fresh12:banking77:ccad56657a290caa7bbe52f6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why is the exchange rate wrong when I purchase something abroad?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0ae3be\": \"The user's request concerns get disposable virtual card.\", \"candidate_c72db8\": \"The user's request concerns transfer fee charged.\", \"candidate_ca839c\": \"The user's request concerns exchange via app.\", \"candidate_b98b5c\": \"The user's request concerns reverted card payment?.\", \"candidate_79e6f9\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_8a8826\": \"The user's request concerns unable to verify identity.\", \"candidate_5f1ace\": \"The user's request concerns top up by cash or cheque.\", \"candidate_3c1357\": \"The user's request concerns contactless not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0ae3be\": 0.0, \"candidate_c72db8\": 0.0, \"candidate_ca839c\": 0.0, \"candidate_b98b5c\": 0.0, \"candidate_79e6f9\": 1.0, \"candidate_8a8826\": 0.0, \"candidate_5f1ace\": 0.0, \"candidate_3c1357\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:fda097511a848874735c99f0:choice", "group_id": "fresh12:banking77:fda097511a848874735c99f0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"You did not apply the correct exchange rate for an item that I bought.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1ba5d2\": \"The user's request concerns country support.\", \"candidate_a4a9ba\": \"The user's request concerns card delivery estimate.\", \"candidate_b64db8\": \"The user's request concerns contactless not working.\", \"candidate_4a8721\": \"The user's request concerns card not working.\", \"candidate_8ae9ec\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_ca88c2\": \"The user's request concerns exchange charge.\", \"candidate_066136\": \"The user's request concerns transfer into account.\", \"candidate_ad3f97\": \"The user's request concerns card payment wrong exchange rate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1ba5d2\": 0.0, \"candidate_a4a9ba\": 0.0, \"candidate_b64db8\": 0.0, \"candidate_4a8721\": 0.0, \"candidate_8ae9ec\": 0.0, \"candidate_ca88c2\": 0.0, \"candidate_066136\": 0.0, \"candidate_ad3f97\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:32561878802bd7783899944b:choice", "group_id": "fresh12:banking77:32561878802bd7783899944b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The rate for a currency exchange was wrong when I bought something.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6be98f\": \"The user's request concerns beneficiary not allowed.\", \"candidate_3265d4\": \"The user's request concerns verify my identity.\", \"candidate_dabdac\": \"The user's request concerns virtual card not working.\", \"candidate_48de82\": \"The user's request concerns compromised card.\", \"candidate_3cffba\": \"The user's request concerns why verify identity.\", \"candidate_e7533a\": \"The user's request concerns card acceptance.\", \"candidate_c077db\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_5e4382\": \"The user's request concerns reverted card payment?.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6be98f\": 0.0, \"candidate_3265d4\": 0.0, \"candidate_dabdac\": 0.0, \"candidate_48de82\": 0.0, \"candidate_3cffba\": 0.0, \"candidate_e7533a\": 0.0, \"candidate_c077db\": 1.0, \"candidate_5e4382\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:30b607a65ae6a0cf0ea950d2:choice", "group_id": "fresh12:banking77:30b607a65ae6a0cf0ea950d2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"the card payment exchange rate is wrong\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_47ddd1\": \"The user's request concerns top up failed.\", \"candidate_ef411b\": \"The user's request concerns verify my identity.\", \"candidate_7fd545\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_447568\": \"The user's request concerns transfer into account.\", \"candidate_2075a0\": \"The user's request concerns age limit.\", \"candidate_1936d8\": \"The user's request concerns pending transfer.\", \"candidate_2cf8d2\": \"The user's request concerns pending card payment.\", \"candidate_475044\": \"The user's request concerns card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_47ddd1\": 0.0, \"candidate_ef411b\": 0.0, \"candidate_7fd545\": 1.0, \"candidate_447568\": 0.0, \"candidate_2075a0\": 0.0, \"candidate_1936d8\": 0.0, \"candidate_2cf8d2\": 0.0, \"candidate_475044\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b5ee6780bb3a61279b291587:choice", "group_id": "fresh12:banking77:b5ee6780bb3a61279b291587", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where did this 1 euro fee come from?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_75473e\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_23bc78\": \"The user's request concerns apple pay or google pay.\", \"candidate_e8149d\": \"The user's request concerns failed transfer.\", \"candidate_e2e5d0\": \"The user's request concerns reverted card payment?.\", \"candidate_ddd511\": \"The user's request concerns verify source of funds.\", \"candidate_671e3b\": \"The user's request concerns extra charge on statement.\", \"candidate_2965eb\": \"The user's request concerns disposable card limits.\", \"candidate_e122ef\": \"The user's request concerns beneficiary not allowed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_75473e\": 0.0, \"candidate_23bc78\": 0.0, \"candidate_e8149d\": 0.0, \"candidate_e2e5d0\": 0.0, \"candidate_ddd511\": 0.0, \"candidate_671e3b\": 1.0, \"candidate_2965eb\": 0.0, \"candidate_e122ef\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:328ba469db251f6507ba5830:choice", "group_id": "fresh12:banking77:328ba469db251f6507ba5830", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I am a new customer, and I happened to look at my app and there is a charge I am not familiar with.  Could you tell me why the extra charge is there?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c0a971\": \"The user's request concerns getting spare card.\", \"candidate_2c7a4b\": \"The user's request concerns topping up by card.\", \"candidate_8e53e5\": \"The user's request concerns card swallowed.\", \"candidate_ec2dd5\": \"The user's request concerns card arrival.\", \"candidate_0e6de8\": \"The user's request concerns extra charge on statement.\", \"candidate_ca8ac5\": \"The user's request concerns apple pay or google pay.\", \"candidate_ce772f\": \"The user's request concerns disposable card limits.\", \"candidate_f67466\": \"The user's request concerns transaction charged twice.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c0a971\": 0.0, \"candidate_2c7a4b\": 0.0, \"candidate_8e53e5\": 0.0, \"candidate_ec2dd5\": 0.0, \"candidate_0e6de8\": 1.0, \"candidate_ca8ac5\": 0.0, \"candidate_ce772f\": 0.0, \"candidate_f67466\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9852fbe6081800e1c6f59dfb:choice", "group_id": "fresh12:banking77:9852fbe6081800e1c6f59dfb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I am seeing an extra fee on my statement what is that for?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1244c4\": \"The user's request concerns extra charge on statement.\", \"candidate_3eed21\": \"The user's request concerns exchange rate.\", \"candidate_b99bb5\": \"The user's request concerns country support.\", \"candidate_8c44c8\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_afb2d4\": \"The user's request concerns declined card payment.\", \"candidate_121334\": \"The user's request concerns age limit.\", \"candidate_9dcce1\": \"The user's request concerns transfer not received by recipient.\", \"candidate_5dde1f\": \"The user's request concerns activate my card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1244c4\": 1.0, \"candidate_3eed21\": 0.0, \"candidate_b99bb5\": 0.0, \"candidate_8c44c8\": 0.0, \"candidate_afb2d4\": 0.0, \"candidate_121334\": 0.0, \"candidate_9dcce1\": 0.0, \"candidate_5dde1f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:58730737b850b9eca077943d:choice", "group_id": "fresh12:banking77:58730737b850b9eca077943d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where did this fee come from?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7e6234\": \"The user's request concerns supported cards and currencies.\", \"candidate_2318cf\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_1bffc1\": \"The user's request concerns declined transfer.\", \"candidate_a997ee\": \"The user's request concerns verify my identity.\", \"candidate_c64316\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_5ef1d8\": \"The user's request concerns extra charge on statement.\", \"candidate_adffaf\": \"The user's request concerns pending cash withdrawal.\", \"candidate_6af47b\": \"The user's request concerns get physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7e6234\": 0.0, \"candidate_2318cf\": 0.0, \"candidate_1bffc1\": 0.0, \"candidate_a997ee\": 0.0, \"candidate_c64316\": 0.0, \"candidate_5ef1d8\": 1.0, \"candidate_adffaf\": 0.0, \"candidate_6af47b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3523a664aba42fea0d36d491:choice", "group_id": "fresh12:banking77:3523a664aba42fea0d36d491", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is an extra charge on my app that I was not told about or I was aware of!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e4e84a\": \"The user's request concerns lost or stolen card.\", \"candidate_b4ae55\": \"The user's request concerns declined cash withdrawal.\", \"candidate_7fca1b\": \"The user's request concerns lost or stolen phone.\", \"candidate_825c9a\": \"The user's request concerns extra charge on statement.\", \"candidate_df1786\": \"The user's request concerns cash withdrawal charge.\", \"candidate_36ca9f\": \"The user's request concerns Refund not showing up.\", \"candidate_85fc5f\": \"The user's request concerns passcode forgotten.\", \"candidate_d483f8\": \"The user's request concerns wrong amount of cash received.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e4e84a\": 0.0, \"candidate_b4ae55\": 0.0, \"candidate_7fca1b\": 0.0, \"candidate_825c9a\": 1.0, \"candidate_df1786\": 0.0, \"candidate_36ca9f\": 0.0, \"candidate_85fc5f\": 0.0, \"candidate_d483f8\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d515a645578cbc2937fc9b2a:choice", "group_id": "fresh12:banking77:d515a645578cbc2937fc9b2a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What would be the reason there's an extra fee on my statement?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_74f70e\": \"The user's request concerns change pin.\", \"candidate_5f80ec\": \"The user's request concerns passcode forgotten.\", \"candidate_3f672f\": \"The user's request concerns virtual card not working.\", \"candidate_0feeef\": \"The user's request concerns transfer timing.\", \"candidate_9188e9\": \"The user's request concerns topping up by card.\", \"candidate_4008ac\": \"The user's request concerns visa or mastercard.\", \"candidate_96bf25\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_d6d009\": \"The user's request concerns extra charge on statement.\"}}}}", "targets_json": "{\"decision\": {\"candidate_74f70e\": 0.0, \"candidate_5f80ec\": 0.0, \"candidate_3f672f\": 0.0, \"candidate_0feeef\": 0.0, \"candidate_9188e9\": 0.0, \"candidate_4008ac\": 0.0, \"candidate_96bf25\": 0.0, \"candidate_d6d009\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a94ea420926197170c1ec2ed:choice", "group_id": "fresh12:banking77:a94ea420926197170c1ec2ed", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long does the cash withdrawal take to pend?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7d1e51\": \"The user's request concerns getting virtual card.\", \"candidate_639c32\": \"The user's request concerns pending cash withdrawal.\", \"candidate_ec9ad8\": \"The user's request concerns verify my identity.\", \"candidate_721539\": \"The user's request concerns card linking.\", \"candidate_04a725\": \"The user's request concerns reverted card payment?.\", \"candidate_fdf07c\": \"The user's request concerns failed transfer.\", \"candidate_bf620c\": \"The user's request concerns fiat currency support.\", \"candidate_3956a8\": \"The user's request concerns wrong exchange rate for cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7d1e51\": 0.0, \"candidate_639c32\": 1.0, \"candidate_ec9ad8\": 0.0, \"candidate_721539\": 0.0, \"candidate_04a725\": 0.0, \"candidate_fdf07c\": 0.0, \"candidate_bf620c\": 0.0, \"candidate_3956a8\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7e9862786b7288dfcb1fcb0e:choice", "group_id": "fresh12:banking77:7e9862786b7288dfcb1fcb0e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What's a pending transaction? My card was declined at an ATM, but my account says it's still pending. Can I cancel the payment?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1cdbf6\": \"The user's request concerns verify my identity.\", \"candidate_9a9423\": \"The user's request concerns receiving money.\", \"candidate_e80d7b\": \"The user's request concerns top up failed.\", \"candidate_29eec7\": \"The user's request concerns pending cash withdrawal.\", \"candidate_38824e\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_5057de\": \"The user's request concerns card payment not recognised.\", \"candidate_53a4a8\": \"The user's request concerns card linking.\", \"candidate_d8d81d\": \"The user's request concerns disposable card limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1cdbf6\": 0.0, \"candidate_9a9423\": 0.0, \"candidate_e80d7b\": 0.0, \"candidate_29eec7\": 1.0, \"candidate_38824e\": 0.0, \"candidate_5057de\": 0.0, \"candidate_53a4a8\": 0.0, \"candidate_d8d81d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:96a9863998f4fdd6ccadab8e:choice", "group_id": "fresh12:banking77:96a9863998f4fdd6ccadab8e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I tried to take money from my card, but it didn't work. Later I saw that the transaction is still in progress. What's goign on?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2ed0d6\": \"The user's request concerns lost or stolen phone.\", \"candidate_8e0f2b\": \"The user's request concerns automatic top up.\", \"candidate_889d08\": \"The user's request concerns failed transfer.\", \"candidate_c5530f\": \"The user's request concerns receiving money.\", \"candidate_af5108\": \"The user's request concerns pending cash withdrawal.\", \"candidate_e18213\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_654687\": \"The user's request concerns transfer into account.\", \"candidate_d79c66\": \"The user's request concerns request refund.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2ed0d6\": 0.0, \"candidate_8e0f2b\": 0.0, \"candidate_889d08\": 0.0, \"candidate_c5530f\": 0.0, \"candidate_af5108\": 1.0, \"candidate_e18213\": 0.0, \"candidate_654687\": 0.0, \"candidate_d79c66\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1e267787a59a597d1d0cb768:choice", "group_id": "fresh12:banking77:1e267787a59a597d1d0cb768", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is my cash withdrawal pending?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7a8eb9\": \"The user's request concerns pending cash withdrawal.\", \"candidate_322711\": \"The user's request concerns terminate account.\", \"candidate_b9f726\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_ad116f\": \"The user's request concerns transfer not received by recipient.\", \"candidate_35ab7a\": \"The user's request concerns verify source of funds.\", \"candidate_1897c2\": \"The user's request concerns order physical card.\", \"candidate_eae0d4\": \"The user's request concerns get disposable virtual card.\", \"candidate_1b5278\": \"The user's request concerns getting spare card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7a8eb9\": 1.0, \"candidate_322711\": 0.0, \"candidate_b9f726\": 0.0, \"candidate_ad116f\": 0.0, \"candidate_35ab7a\": 0.0, \"candidate_1897c2\": 0.0, \"candidate_eae0d4\": 0.0, \"candidate_1b5278\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e107be55b6c122203b3b9063:choice", "group_id": "fresh12:banking77:e107be55b6c122203b3b9063", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I got some cash of an ATM earlier but this shows up as pending in the app. How can this be still pending, I've got the cash already?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_108d0d\": \"The user's request concerns apple pay or google pay.\", \"candidate_99859c\": \"The user's request concerns declined transfer.\", \"candidate_bf186a\": \"The user's request concerns failed transfer.\", \"candidate_91dae0\": \"The user's request concerns pending cash withdrawal.\", \"candidate_d77f57\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_ba8cf9\": \"The user's request concerns pending transfer.\", \"candidate_3ec526\": \"The user's request concerns virtual card not working.\", \"candidate_457c36\": \"The user's request concerns edit personal details.\"}}}}", "targets_json": "{\"decision\": {\"candidate_108d0d\": 0.0, \"candidate_99859c\": 0.0, \"candidate_bf186a\": 0.0, \"candidate_91dae0\": 1.0, \"candidate_d77f57\": 0.0, \"candidate_ba8cf9\": 0.0, \"candidate_3ec526\": 0.0, \"candidate_457c36\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:edd1d0af1cf1519ad49a8063:choice", "group_id": "fresh12:banking77:edd1d0af1cf1519ad49a8063", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I'm not quite understanding why my cash withdrawal is showing as pending.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_df011f\": \"The user's request concerns top up by cash or cheque.\", \"candidate_b70b3d\": \"The user's request concerns pending cash withdrawal.\", \"candidate_6a0a24\": \"The user's request concerns card swallowed.\", \"candidate_2a0388\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_b820ac\": \"The user's request concerns fiat currency support.\", \"candidate_3f0922\": \"The user's request concerns lost or stolen phone.\", \"candidate_dde2da\": \"The user's request concerns activate my card.\", \"candidate_05942e\": \"The user's request concerns why verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_df011f\": 0.0, \"candidate_b70b3d\": 1.0, \"candidate_6a0a24\": 0.0, \"candidate_2a0388\": 0.0, \"candidate_b820ac\": 0.0, \"candidate_3f0922\": 0.0, \"candidate_dde2da\": 0.0, \"candidate_05942e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:79d6f0f28e804c2405c79c23:choice", "group_id": "fresh12:banking77:79d6f0f28e804c2405c79c23", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Do you accept exchanges to EUR?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8e41a9\": \"The user's request concerns atm support.\", \"candidate_f7ed25\": \"The user's request concerns visa or mastercard.\", \"candidate_0f4e8c\": \"The user's request concerns fiat currency support.\", \"candidate_f6ae0c\": \"The user's request concerns cancel transfer.\", \"candidate_a3118e\": \"The user's request concerns transaction charged twice.\", \"candidate_430170\": \"The user's request concerns verify my identity.\", \"candidate_c5920a\": \"The user's request concerns card about to expire.\", \"candidate_6ae7e0\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8e41a9\": 0.0, \"candidate_f7ed25\": 0.0, \"candidate_0f4e8c\": 1.0, \"candidate_f6ae0c\": 0.0, \"candidate_a3118e\": 0.0, \"candidate_430170\": 0.0, \"candidate_c5920a\": 0.0, \"candidate_6ae7e0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:148335257b980b3d1841b7e5:choice", "group_id": "fresh12:banking77:148335257b980b3d1841b7e5", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What kind of currency can I hold money in?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_aacde0\": \"The user's request concerns fiat currency support.\", \"candidate_242ad9\": \"The user's request concerns card about to expire.\", \"candidate_8c3d6f\": \"The user's request concerns contactless not working.\", \"candidate_b5644d\": \"The user's request concerns disposable card limits.\", \"candidate_bf9488\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_445573\": \"The user's request concerns card not working.\", \"candidate_6ad8c9\": \"The user's request concerns card linking.\", \"candidate_c147a2\": \"The user's request concerns apple pay or google pay.\"}}}}", "targets_json": "{\"decision\": {\"candidate_aacde0\": 1.0, \"candidate_242ad9\": 0.0, \"candidate_8c3d6f\": 0.0, \"candidate_b5644d\": 0.0, \"candidate_bf9488\": 0.0, \"candidate_445573\": 0.0, \"candidate_6ad8c9\": 0.0, \"candidate_c147a2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:30f7966dfee901b6a1be0314:choice", "group_id": "fresh12:banking77:30f7966dfee901b6a1be0314", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can someone assist me with the details explaining which fiat currencies that you support?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_589413\": \"The user's request concerns apple pay or google pay.\", \"candidate_99fa89\": \"The user's request concerns failed transfer.\", \"candidate_69e9a5\": \"The user's request concerns disposable card limits.\", \"candidate_968a4f\": \"The user's request concerns transfer fee charged.\", \"candidate_eed0d1\": \"The user's request concerns supported cards and currencies.\", \"candidate_1c8d6f\": \"The user's request concerns top up reverted.\", \"candidate_f9c842\": \"The user's request concerns card arrival.\", \"candidate_d91138\": \"The user's request concerns fiat currency support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_589413\": 0.0, \"candidate_99fa89\": 0.0, \"candidate_69e9a5\": 0.0, \"candidate_968a4f\": 0.0, \"candidate_eed0d1\": 0.0, \"candidate_1c8d6f\": 0.0, \"candidate_f9c842\": 0.0, \"candidate_d91138\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:59c61574717eb3369a1d8e4a:choice", "group_id": "fresh12:banking77:59c61574717eb3369a1d8e4a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is it permissible to hold money in multiple currencies?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e6bc28\": \"The user's request concerns getting virtual card.\", \"candidate_61ec89\": \"The user's request concerns declined card payment.\", \"candidate_c50946\": \"The user's request concerns pending top up.\", \"candidate_639dee\": \"The user's request concerns transfer not received by recipient.\", \"candidate_447a5c\": \"The user's request concerns fiat currency support.\", \"candidate_5e3380\": \"The user's request concerns disposable card limits.\", \"candidate_d62f43\": \"The user's request concerns get physical card.\", \"candidate_313eb4\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e6bc28\": 0.0, \"candidate_61ec89\": 0.0, \"candidate_c50946\": 0.0, \"candidate_639dee\": 0.0, \"candidate_447a5c\": 1.0, \"candidate_5e3380\": 0.0, \"candidate_d62f43\": 0.0, \"candidate_313eb4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:351a7769323e7bf01ae49793:choice", "group_id": "fresh12:banking77:351a7769323e7bf01ae49793", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Besides USD what other currencies can I have?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3d381c\": \"The user's request concerns why verify identity.\", \"candidate_4ee312\": \"The user's request concerns fiat currency support.\", \"candidate_acbf6a\": \"The user's request concerns transfer into account.\", \"candidate_b2f42b\": \"The user's request concerns card swallowed.\", \"candidate_8eaa8c\": \"The user's request concerns getting spare card.\", \"candidate_c4bd85\": \"The user's request concerns terminate account.\", \"candidate_921b84\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_913d82\": \"The user's request concerns supported cards and currencies.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3d381c\": 0.0, \"candidate_4ee312\": 1.0, \"candidate_acbf6a\": 0.0, \"candidate_b2f42b\": 0.0, \"candidate_8eaa8c\": 0.0, \"candidate_c4bd85\": 0.0, \"candidate_921b84\": 0.0, \"candidate_913d82\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:485c598389097e02240d343f:choice", "group_id": "fresh12:banking77:485c598389097e02240d343f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What countries can I travel to and have access to their currencies?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_695109\": \"The user's request concerns extra charge on statement.\", \"candidate_1df594\": \"The user's request concerns fiat currency support.\", \"candidate_c43637\": \"The user's request concerns card payment not recognised.\", \"candidate_5f80f0\": \"The user's request concerns failed transfer.\", \"candidate_42abc5\": \"The user's request concerns top up reverted.\", \"candidate_b672ec\": \"The user's request concerns terminate account.\", \"candidate_f3f1f5\": \"The user's request concerns apple pay or google pay.\", \"candidate_fc5531\": \"The user's request concerns Refund not showing up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_695109\": 0.0, \"candidate_1df594\": 1.0, \"candidate_c43637\": 0.0, \"candidate_5f80f0\": 0.0, \"candidate_42abc5\": 0.0, \"candidate_b672ec\": 0.0, \"candidate_f3f1f5\": 0.0, \"candidate_fc5531\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:680d2a628cf3f40de526a20d:choice", "group_id": "fresh12:banking77:680d2a628cf3f40de526a20d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need my card quickly\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2468d8\": \"The user's request concerns order physical card.\", \"candidate_6bec26\": \"The user's request concerns card delivery estimate.\", \"candidate_d82c01\": \"The user's request concerns pending cash withdrawal.\", \"candidate_c5b60f\": \"The user's request concerns card linking.\", \"candidate_7a02cb\": \"The user's request concerns exchange rate.\", \"candidate_1399b2\": \"The user's request concerns get disposable virtual card.\", \"candidate_743e7c\": \"The user's request concerns disposable card limits.\", \"candidate_5c2b4c\": \"The user's request concerns wrong amount of cash received.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2468d8\": 0.0, \"candidate_6bec26\": 1.0, \"candidate_d82c01\": 0.0, \"candidate_c5b60f\": 0.0, \"candidate_7a02cb\": 0.0, \"candidate_1399b2\": 0.0, \"candidate_743e7c\": 0.0, \"candidate_5c2b4c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4745797c35219f6e3d47c773:choice", "group_id": "fresh12:banking77:4745797c35219f6e3d47c773", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What's the U.S. Delivery time?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d23674\": \"The user's request concerns get physical card.\", \"candidate_4a5651\": \"The user's request concerns edit personal details.\", \"candidate_81f1fc\": \"The user's request concerns apple pay or google pay.\", \"candidate_288080\": \"The user's request concerns lost or stolen phone.\", \"candidate_73b243\": \"The user's request concerns card not working.\", \"candidate_761b07\": \"The user's request concerns card delivery estimate.\", \"candidate_868d68\": \"The user's request concerns wrong amount of cash received.\", \"candidate_705f92\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d23674\": 0.0, \"candidate_4a5651\": 0.0, \"candidate_81f1fc\": 0.0, \"candidate_288080\": 0.0, \"candidate_73b243\": 0.0, \"candidate_761b07\": 1.0, \"candidate_868d68\": 0.0, \"candidate_705f92\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:af3be12a2a47a1304e9e0c4d:choice", "group_id": "fresh12:banking77:af3be12a2a47a1304e9e0c4d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long will it take to get to me?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2cd06d\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_dcb6a3\": \"The user's request concerns card acceptance.\", \"candidate_a34640\": \"The user's request concerns card delivery estimate.\", \"candidate_7cf750\": \"The user's request concerns order physical card.\", \"candidate_abd0cf\": \"The user's request concerns get physical card.\", \"candidate_abc3d4\": \"The user's request concerns card payment fee charged.\", \"candidate_bfa110\": \"The user's request concerns pending cash withdrawal.\", \"candidate_9f391d\": \"The user's request concerns reverted card payment?.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2cd06d\": 0.0, \"candidate_dcb6a3\": 0.0, \"candidate_a34640\": 1.0, \"candidate_7cf750\": 0.0, \"candidate_abd0cf\": 0.0, \"candidate_abc3d4\": 0.0, \"candidate_bfa110\": 0.0, \"candidate_9f391d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2fed86e3ed04f6fa3235f615:choice", "group_id": "fresh12:banking77:2fed86e3ed04f6fa3235f615", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Ordered awhile back, what is the ETA in the US?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4f487d\": \"The user's request concerns disposable card limits.\", \"candidate_ec49e5\": \"The user's request concerns transfer fee charged.\", \"candidate_e1564a\": \"The user's request concerns card swallowed.\", \"candidate_be44c4\": \"The user's request concerns change pin.\", \"candidate_069a43\": \"The user's request concerns card delivery estimate.\", \"candidate_85310b\": \"The user's request concerns declined card payment.\", \"candidate_4a98fa\": \"The user's request concerns transfer into account.\", \"candidate_282c80\": \"The user's request concerns transaction charged twice.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4f487d\": 0.0, \"candidate_ec49e5\": 0.0, \"candidate_e1564a\": 0.0, \"candidate_be44c4\": 0.0, \"candidate_069a43\": 1.0, \"candidate_85310b\": 0.0, \"candidate_4a98fa\": 0.0, \"candidate_282c80\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:fac0492013b8a7c7a3c5b755:choice", "group_id": "fresh12:banking77:fac0492013b8a7c7a3c5b755", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I make sure my card is delivered on a specific day?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_066a94\": \"The user's request concerns card delivery estimate.\", \"candidate_8ddc98\": \"The user's request concerns transfer not received by recipient.\", \"candidate_be3dac\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_9b4229\": \"The user's request concerns pending transfer.\", \"candidate_a20688\": \"The user's request concerns passcode forgotten.\", \"candidate_443370\": \"The user's request concerns transfer timing.\", \"candidate_2aa901\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_052c3b\": \"The user's request concerns declined card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_066a94\": 1.0, \"candidate_8ddc98\": 0.0, \"candidate_be3dac\": 0.0, \"candidate_9b4229\": 0.0, \"candidate_a20688\": 0.0, \"candidate_443370\": 0.0, \"candidate_2aa901\": 0.0, \"candidate_052c3b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4f75fd1f7effddf514f769e4:choice", "group_id": "fresh12:banking77:4f75fd1f7effddf514f769e4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need it delivered on a certain date.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9df0ae\": \"The user's request concerns exchange via app.\", \"candidate_1eea08\": \"The user's request concerns pending top up.\", \"candidate_a16344\": \"The user's request concerns compromised card.\", \"candidate_c08643\": \"The user's request concerns age limit.\", \"candidate_cc3f90\": \"The user's request concerns card delivery estimate.\", \"candidate_01c80c\": \"The user's request concerns terminate account.\", \"candidate_d32df7\": \"The user's request concerns declined cash withdrawal.\", \"candidate_5b85d0\": \"The user's request concerns lost or stolen card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9df0ae\": 0.0, \"candidate_1eea08\": 0.0, \"candidate_a16344\": 0.0, \"candidate_c08643\": 0.0, \"candidate_cc3f90\": 1.0, \"candidate_01c80c\": 0.0, \"candidate_d32df7\": 0.0, \"candidate_5b85d0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:47681679de462643295dda35:choice", "group_id": "fresh12:banking77:47681679de462643295dda35", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I enable auto-top up?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2cf6d5\": \"The user's request concerns exchange charge.\", \"candidate_347c80\": \"The user's request concerns card delivery estimate.\", \"candidate_c91786\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_43b22e\": \"The user's request concerns fiat currency support.\", \"candidate_d832e4\": \"The user's request concerns pending transfer.\", \"candidate_e819c1\": \"The user's request concerns why verify identity.\", \"candidate_af047b\": \"The user's request concerns automatic top up.\", \"candidate_9cf570\": \"The user's request concerns cancel transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2cf6d5\": 0.0, \"candidate_347c80\": 0.0, \"candidate_c91786\": 0.0, \"candidate_43b22e\": 0.0, \"candidate_d832e4\": 0.0, \"candidate_e819c1\": 0.0, \"candidate_af047b\": 1.0, \"candidate_9cf570\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:681687b6be17f25cf749bb30:choice", "group_id": "fresh12:banking77:681687b6be17f25cf749bb30", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I space out how often money is transferred along my travels?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_23b2a9\": \"The user's request concerns card payment not recognised.\", \"candidate_7f8c97\": \"The user's request concerns pending top up.\", \"candidate_998491\": \"The user's request concerns transaction charged twice.\", \"candidate_d6266a\": \"The user's request concerns wrong amount of cash received.\", \"candidate_a86129\": \"The user's request concerns Refund not showing up.\", \"candidate_aedf65\": \"The user's request concerns automatic top up.\", \"candidate_f8198f\": \"The user's request concerns receiving money.\", \"candidate_7fa757\": \"The user's request concerns exchange via app.\"}}}}", "targets_json": "{\"decision\": {\"candidate_23b2a9\": 0.0, \"candidate_7f8c97\": 0.0, \"candidate_998491\": 0.0, \"candidate_d6266a\": 0.0, \"candidate_a86129\": 0.0, \"candidate_aedf65\": 1.0, \"candidate_f8198f\": 0.0, \"candidate_7fa757\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4050b9c734c47989f830bde7:choice", "group_id": "fresh12:banking77:4050b9c734c47989f830bde7", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can it automatically top-up money if there isn't much left?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_44f1b9\": \"The user's request concerns wrong amount of cash received.\", \"candidate_e38bc4\": \"The user's request concerns transfer timing.\", \"candidate_63f0de\": \"The user's request concerns getting virtual card.\", \"candidate_c27e04\": \"The user's request concerns getting spare card.\", \"candidate_8e1321\": \"The user's request concerns card about to expire.\", \"candidate_2fbe61\": \"The user's request concerns top up reverted.\", \"candidate_649c9d\": \"The user's request concerns automatic top up.\", \"candidate_e6ea03\": \"The user's request concerns transfer not received by recipient.\"}}}}", "targets_json": "{\"decision\": {\"candidate_44f1b9\": 0.0, \"candidate_e38bc4\": 0.0, \"candidate_63f0de\": 0.0, \"candidate_c27e04\": 0.0, \"candidate_8e1321\": 0.0, \"candidate_2fbe61\": 0.0, \"candidate_649c9d\": 1.0, \"candidate_e6ea03\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:df97efc71d42300d14751787:choice", "group_id": "fresh12:banking77:df97efc71d42300d14751787", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where is the auto top-up option?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f88212\": \"The user's request concerns atm support.\", \"candidate_8b3db2\": \"The user's request concerns unable to verify identity.\", \"candidate_9b9905\": \"The user's request concerns request refund.\", \"candidate_c155a7\": \"The user's request concerns visa or mastercard.\", \"candidate_6035ed\": \"The user's request concerns automatic top up.\", \"candidate_9c46d2\": \"The user's request concerns wrong amount of cash received.\", \"candidate_f70250\": \"The user's request concerns transfer not received by recipient.\", \"candidate_5a1d8a\": \"The user's request concerns top up by bank transfer charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f88212\": 0.0, \"candidate_8b3db2\": 0.0, \"candidate_9b9905\": 0.0, \"candidate_c155a7\": 0.0, \"candidate_6035ed\": 1.0, \"candidate_9c46d2\": 0.0, \"candidate_f70250\": 0.0, \"candidate_5a1d8a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:cadc633c441a69536bebbf04:choice", "group_id": "fresh12:banking77:cadc633c441a69536bebbf04", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I have it add money at certain time intervals when I travel?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ffcab2\": \"The user's request concerns transfer not received by recipient.\", \"candidate_bcda44\": \"The user's request concerns lost or stolen phone.\", \"candidate_c63dae\": \"The user's request concerns transfer timing.\", \"candidate_29723f\": \"The user's request concerns transfer fee charged.\", \"candidate_50b3ee\": \"The user's request concerns automatic top up.\", \"candidate_c6526c\": \"The user's request concerns activate my card.\", \"candidate_b6c6e5\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_502733\": \"The user's request concerns card arrival.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ffcab2\": 0.0, \"candidate_bcda44\": 0.0, \"candidate_c63dae\": 0.0, \"candidate_29723f\": 0.0, \"candidate_50b3ee\": 1.0, \"candidate_c6526c\": 0.0, \"candidate_b6c6e5\": 0.0, \"candidate_502733\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5d9547ecd1d962f01002daeb:choice", "group_id": "fresh12:banking77:5d9547ecd1d962f01002daeb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a top-up feature, if there isn't a lot of money left?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_53e5a7\": \"The user's request concerns card linking.\", \"candidate_4da414\": \"The user's request concerns age limit.\", \"candidate_b360c2\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_f857a2\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_08a387\": \"The user's request concerns top up failed.\", \"candidate_ade6ac\": \"The user's request concerns automatic top up.\", \"candidate_e2cd83\": \"The user's request concerns exchange via app.\", \"candidate_2dc509\": \"The user's request concerns get disposable virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_53e5a7\": 0.0, \"candidate_4da414\": 0.0, \"candidate_b360c2\": 0.0, \"candidate_f857a2\": 0.0, \"candidate_08a387\": 0.0, \"candidate_ade6ac\": 1.0, \"candidate_e2cd83\": 0.0, \"candidate_2dc509\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:71e966029a6d41b123e4fc9b:choice", "group_id": "fresh12:banking77:71e966029a6d41b123e4fc9b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can I resolve a problem where my card won't go through at all?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_32dd74\": \"The user's request concerns automatic top up.\", \"candidate_ca5927\": \"The user's request concerns get disposable virtual card.\", \"candidate_04e6b9\": \"The user's request concerns card not working.\", \"candidate_6d8aac\": \"The user's request concerns pending top up.\", \"candidate_b7a428\": \"The user's request concerns cancel transfer.\", \"candidate_d9773e\": \"The user's request concerns request refund.\", \"candidate_fe8e3b\": \"The user's request concerns country support.\", \"candidate_b808e7\": \"The user's request concerns contactless not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_32dd74\": 0.0, \"candidate_ca5927\": 0.0, \"candidate_04e6b9\": 1.0, \"candidate_6d8aac\": 0.0, \"candidate_b7a428\": 0.0, \"candidate_d9773e\": 0.0, \"candidate_fe8e3b\": 0.0, \"candidate_b808e7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:cabf2d8f8c6adab0563f3c68:choice", "group_id": "fresh12:banking77:cabf2d8f8c6adab0563f3c68", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My actual card isn't working.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_25300a\": \"The user's request concerns Refund not showing up.\", \"candidate_230946\": \"The user's request concerns pending cash withdrawal.\", \"candidate_7f8ad3\": \"The user's request concerns card payment not recognised.\", \"candidate_4c5275\": \"The user's request concerns visa or mastercard.\", \"candidate_f0a1d0\": \"The user's request concerns beneficiary not allowed.\", \"candidate_1754a0\": \"The user's request concerns transfer not received by recipient.\", \"candidate_c9c354\": \"The user's request concerns pending top up.\", \"candidate_c9393e\": \"The user's request concerns card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_25300a\": 0.0, \"candidate_230946\": 0.0, \"candidate_7f8ad3\": 0.0, \"candidate_4c5275\": 0.0, \"candidate_f0a1d0\": 0.0, \"candidate_1754a0\": 0.0, \"candidate_c9c354\": 0.0, \"candidate_c9393e\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:da95941bbf32e74164bd913d:choice", "group_id": "fresh12:banking77:da95941bbf32e74164bd913d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I've tried my card a bunch of times and it never worked.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_951255\": \"The user's request concerns card arrival.\", \"candidate_9b3dbb\": \"The user's request concerns beneficiary not allowed.\", \"candidate_b54469\": \"The user's request concerns get disposable virtual card.\", \"candidate_df7b0d\": \"The user's request concerns fiat currency support.\", \"candidate_dfc1f9\": \"The user's request concerns top up limits.\", \"candidate_f93f7d\": \"The user's request concerns card not working.\", \"candidate_503e1e\": \"The user's request concerns verify source of funds.\", \"candidate_0f527a\": \"The user's request concerns exchange via app.\"}}}}", "targets_json": "{\"decision\": {\"candidate_951255\": 0.0, \"candidate_9b3dbb\": 0.0, \"candidate_b54469\": 0.0, \"candidate_df7b0d\": 0.0, \"candidate_dfc1f9\": 0.0, \"candidate_f93f7d\": 1.0, \"candidate_503e1e\": 0.0, \"candidate_0f527a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ce679af1b02c4697becd2ae8:choice", "group_id": "fresh12:banking77:ce679af1b02c4697becd2ae8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card broke.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d61526\": \"The user's request concerns exchange rate.\", \"candidate_b95faa\": \"The user's request concerns pin blocked.\", \"candidate_1a60f6\": \"The user's request concerns atm support.\", \"candidate_fa013f\": \"The user's request concerns declined card payment.\", \"candidate_a0aeee\": \"The user's request concerns top up by card charge.\", \"candidate_d6344d\": \"The user's request concerns card not working.\", \"candidate_d0969e\": \"The user's request concerns activate my card.\", \"candidate_a2d143\": \"The user's request concerns verify my identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d61526\": 0.0, \"candidate_b95faa\": 0.0, \"candidate_1a60f6\": 0.0, \"candidate_fa013f\": 0.0, \"candidate_a0aeee\": 0.0, \"candidate_d6344d\": 1.0, \"candidate_d0969e\": 0.0, \"candidate_a2d143\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b05d3d93ff9777925bf942e0:choice", "group_id": "fresh12:banking77:b05d3d93ff9777925bf942e0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I find out why my card won't work?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_46fdc8\": \"The user's request concerns top up failed.\", \"candidate_5375d7\": \"The user's request concerns transfer not received by recipient.\", \"candidate_53cc9a\": \"The user's request concerns card not working.\", \"candidate_2480d3\": \"The user's request concerns disposable card limits.\", \"candidate_8eaf8a\": \"The user's request concerns exchange via app.\", \"candidate_c2a782\": \"The user's request concerns pending cash withdrawal.\", \"candidate_de3d82\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_cbc4d6\": \"The user's request concerns card payment wrong exchange rate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_46fdc8\": 0.0, \"candidate_5375d7\": 0.0, \"candidate_53cc9a\": 1.0, \"candidate_2480d3\": 0.0, \"candidate_8eaf8a\": 0.0, \"candidate_c2a782\": 0.0, \"candidate_de3d82\": 0.0, \"candidate_cbc4d6\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ff9a9d3acf5b9b68349ef8ee:choice", "group_id": "fresh12:banking77:ff9a9d3acf5b9b68349ef8ee", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I cannot use my card.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dbaba8\": \"The user's request concerns fiat currency support.\", \"candidate_2de657\": \"The user's request concerns exchange via app.\", \"candidate_af8be2\": \"The user's request concerns card payment not recognised.\", \"candidate_061c8e\": \"The user's request concerns card linking.\", \"candidate_a71b3d\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_8c0c82\": \"The user's request concerns card not working.\", \"candidate_4f7334\": \"The user's request concerns failed transfer.\", \"candidate_3053ef\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dbaba8\": 0.0, \"candidate_2de657\": 0.0, \"candidate_af8be2\": 0.0, \"candidate_061c8e\": 0.0, \"candidate_a71b3d\": 0.0, \"candidate_8c0c82\": 1.0, \"candidate_4f7334\": 0.0, \"candidate_3053ef\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3f1cc7fb016ad7e7118a4a12:choice", "group_id": "fresh12:banking77:3f1cc7fb016ad7e7118a4a12", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I change from one currency to another?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_422cb3\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_9cf1b7\": \"The user's request concerns cancel transfer.\", \"candidate_a5c580\": \"The user's request concerns pin blocked.\", \"candidate_ceeda3\": \"The user's request concerns exchange via app.\", \"candidate_d398fe\": \"The user's request concerns cash withdrawal charge.\", \"candidate_23e175\": \"The user's request concerns change pin.\", \"candidate_a47a8d\": \"The user's request concerns virtual card not working.\", \"candidate_30c6ec\": \"The user's request concerns top up limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_422cb3\": 0.0, \"candidate_9cf1b7\": 0.0, \"candidate_a5c580\": 0.0, \"candidate_ceeda3\": 1.0, \"candidate_d398fe\": 0.0, \"candidate_23e175\": 0.0, \"candidate_a47a8d\": 0.0, \"candidate_30c6ec\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f2ff774ee5bc6e37f87fb084:choice", "group_id": "fresh12:banking77:f2ff774ee5bc6e37f87fb084", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I was wondering whether exchanging currencies can be done on this app.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b3ee80\": \"The user's request concerns request refund.\", \"candidate_9e1a35\": \"The user's request concerns getting virtual card.\", \"candidate_6999fb\": \"The user's request concerns card payment not recognised.\", \"candidate_3482b1\": \"The user's request concerns card about to expire.\", \"candidate_7a1732\": \"The user's request concerns apple pay or google pay.\", \"candidate_41870f\": \"The user's request concerns top up failed.\", \"candidate_4db995\": \"The user's request concerns exchange via app.\", \"candidate_306c14\": \"The user's request concerns pending transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b3ee80\": 0.0, \"candidate_9e1a35\": 0.0, \"candidate_6999fb\": 0.0, \"candidate_3482b1\": 0.0, \"candidate_7a1732\": 0.0, \"candidate_41870f\": 0.0, \"candidate_4db995\": 1.0, \"candidate_306c14\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:011df98e5f8219ad4344252b:choice", "group_id": "fresh12:banking77:011df98e5f8219ad4344252b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I interconvert between AUD and GBP?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bc9354\": \"The user's request concerns card acceptance.\", \"candidate_8bb9f2\": \"The user's request concerns Refund not showing up.\", \"candidate_752bc1\": \"The user's request concerns why verify identity.\", \"candidate_37bfc1\": \"The user's request concerns exchange via app.\", \"candidate_04cb1b\": \"The user's request concerns cash withdrawal charge.\", \"candidate_e8b89f\": \"The user's request concerns declined transfer.\", \"candidate_e46fcd\": \"The user's request concerns failed transfer.\", \"candidate_880cc7\": \"The user's request concerns exchange charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bc9354\": 0.0, \"candidate_8bb9f2\": 0.0, \"candidate_752bc1\": 0.0, \"candidate_37bfc1\": 1.0, \"candidate_04cb1b\": 0.0, \"candidate_e8b89f\": 0.0, \"candidate_e46fcd\": 0.0, \"candidate_880cc7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d8553852434fd47ac4a895ae:choice", "group_id": "fresh12:banking77:d8553852434fd47ac4a895ae", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"If I receive foreign currencies, am I able to exchange them on the app?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d212a1\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_38d4e1\": \"The user's request concerns transaction charged twice.\", \"candidate_cede94\": \"The user's request concerns request refund.\", \"candidate_fadbbb\": \"The user's request concerns exchange via app.\", \"candidate_8b4d64\": \"The user's request concerns top up reverted.\", \"candidate_eac6bd\": \"The user's request concerns visa or mastercard.\", \"candidate_dad1ed\": \"The user's request concerns transfer not received by recipient.\", \"candidate_1e3915\": \"The user's request concerns contactless not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d212a1\": 0.0, \"candidate_38d4e1\": 0.0, \"candidate_cede94\": 0.0, \"candidate_fadbbb\": 1.0, \"candidate_8b4d64\": 0.0, \"candidate_eac6bd\": 0.0, \"candidate_dad1ed\": 0.0, \"candidate_1e3915\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:eeb1302aa5292b6c7694c776:choice", "group_id": "fresh12:banking77:eeb1302aa5292b6c7694c776", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How are currencies exchanged?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2c6ac4\": \"The user's request concerns verify source of funds.\", \"candidate_08c770\": \"The user's request concerns virtual card not working.\", \"candidate_8f732c\": \"The user's request concerns exchange charge.\", \"candidate_1937a9\": \"The user's request concerns exchange via app.\", \"candidate_b3bd99\": \"The user's request concerns card delivery estimate.\", \"candidate_9a042e\": \"The user's request concerns pending cash withdrawal.\", \"candidate_e308e6\": \"The user's request concerns transfer into account.\", \"candidate_2f251b\": \"The user's request concerns failed transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2c6ac4\": 0.0, \"candidate_08c770\": 0.0, \"candidate_8f732c\": 0.0, \"candidate_1937a9\": 1.0, \"candidate_b3bd99\": 0.0, \"candidate_9a042e\": 0.0, \"candidate_e308e6\": 0.0, \"candidate_2f251b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8b9db084852d55e69153c20d:choice", "group_id": "fresh12:banking77:8b9db084852d55e69153c20d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How would I use my money in a different country?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_025233\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_d1e655\": \"The user's request concerns pending top up.\", \"candidate_5e34f1\": \"The user's request concerns exchange via app.\", \"candidate_bc7223\": \"The user's request concerns card acceptance.\", \"candidate_701079\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_fb4a69\": \"The user's request concerns top up by card charge.\", \"candidate_3a5749\": \"The user's request concerns terminate account.\", \"candidate_693059\": \"The user's request concerns country support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_025233\": 0.0, \"candidate_d1e655\": 0.0, \"candidate_5e34f1\": 1.0, \"candidate_bc7223\": 0.0, \"candidate_701079\": 0.0, \"candidate_fb4a69\": 0.0, \"candidate_3a5749\": 0.0, \"candidate_693059\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9fbb050d0e54b72002145d88:choice", "group_id": "fresh12:banking77:9fbb050d0e54b72002145d88", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Someone stole my card. I need to report it stolen. I made a police report already, but how do I report it with you?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f3783f\": \"The user's request concerns transfer not received by recipient.\", \"candidate_dd4841\": \"The user's request concerns supported cards and currencies.\", \"candidate_6a169f\": \"The user's request concerns fiat currency support.\", \"candidate_4a7818\": \"The user's request concerns card linking.\", \"candidate_7e9a0d\": \"The user's request concerns lost or stolen card.\", \"candidate_fed433\": \"The user's request concerns top up failed.\", \"candidate_3bae49\": \"The user's request concerns transfer into account.\", \"candidate_58573c\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f3783f\": 0.0, \"candidate_dd4841\": 0.0, \"candidate_6a169f\": 0.0, \"candidate_4a7818\": 0.0, \"candidate_7e9a0d\": 1.0, \"candidate_fed433\": 0.0, \"candidate_3bae49\": 0.0, \"candidate_58573c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:cd22503b5eadf31f948d6825:choice", "group_id": "fresh12:banking77:cd22503b5eadf31f948d6825", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need help with a lost card\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cfafff\": \"The user's request concerns top up by card charge.\", \"candidate_f25a3c\": \"The user's request concerns exchange rate.\", \"candidate_c7a875\": \"The user's request concerns request refund.\", \"candidate_b0c3c0\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_977914\": \"The user's request concerns contactless not working.\", \"candidate_47599a\": \"The user's request concerns cancel transfer.\", \"candidate_79a669\": \"The user's request concerns lost or stolen card.\", \"candidate_0b04d0\": \"The user's request concerns virtual card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cfafff\": 0.0, \"candidate_f25a3c\": 0.0, \"candidate_c7a875\": 0.0, \"candidate_b0c3c0\": 0.0, \"candidate_977914\": 0.0, \"candidate_47599a\": 0.0, \"candidate_79a669\": 1.0, \"candidate_0b04d0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b0158cc1fa5b0f321e23dfce:choice", "group_id": "fresh12:banking77:b0158cc1fa5b0f321e23dfce", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is a copy of the police report necessary for completing the report process?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_762307\": \"The user's request concerns lost or stolen card.\", \"candidate_9d73b2\": \"The user's request concerns pending cash withdrawal.\", \"candidate_281817\": \"The user's request concerns card arrival.\", \"candidate_f11bfa\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_6c797f\": \"The user's request concerns reverted card payment?.\", \"candidate_81f6a9\": \"The user's request concerns declined transfer.\", \"candidate_425634\": \"The user's request concerns compromised card.\", \"candidate_bc2a8b\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_762307\": 1.0, \"candidate_9d73b2\": 0.0, \"candidate_281817\": 0.0, \"candidate_f11bfa\": 0.0, \"candidate_6c797f\": 0.0, \"candidate_81f6a9\": 0.0, \"candidate_425634\": 0.0, \"candidate_bc2a8b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e46c8584b5b2aa8cef1ab0bc:choice", "group_id": "fresh12:banking77:e46c8584b5b2aa8cef1ab0bc", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card was taken from me\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_60db07\": \"The user's request concerns exchange via app.\", \"candidate_10b1bf\": \"The user's request concerns request refund.\", \"candidate_ec62c9\": \"The user's request concerns receiving money.\", \"candidate_f0807f\": \"The user's request concerns transfer fee charged.\", \"candidate_0dde0a\": \"The user's request concerns edit personal details.\", \"candidate_d417ba\": \"The user's request concerns getting spare card.\", \"candidate_be3147\": \"The user's request concerns card not working.\", \"candidate_8547ca\": \"The user's request concerns lost or stolen card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_60db07\": 0.0, \"candidate_10b1bf\": 0.0, \"candidate_ec62c9\": 0.0, \"candidate_f0807f\": 0.0, \"candidate_0dde0a\": 0.0, \"candidate_d417ba\": 0.0, \"candidate_be3147\": 0.0, \"candidate_8547ca\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0e3fe69e785c0da5f4b1f9d6:choice", "group_id": "fresh12:banking77:0e3fe69e785c0da5f4b1f9d6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card is lost! What do I do now?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_627184\": \"The user's request concerns wrong amount of cash received.\", \"candidate_514264\": \"The user's request concerns lost or stolen card.\", \"candidate_9b02c1\": \"The user's request concerns exchange charge.\", \"candidate_040ec7\": \"The user's request concerns virtual card not working.\", \"candidate_a3e52b\": \"The user's request concerns visa or mastercard.\", \"candidate_5a6e7a\": \"The user's request concerns pending transfer.\", \"candidate_514ffc\": \"The user's request concerns exchange via app.\", \"candidate_8d2a54\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_627184\": 0.0, \"candidate_514264\": 1.0, \"candidate_9b02c1\": 0.0, \"candidate_040ec7\": 0.0, \"candidate_a3e52b\": 0.0, \"candidate_5a6e7a\": 0.0, \"candidate_514ffc\": 0.0, \"candidate_8d2a54\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c851b164c70a9c361826934b:choice", "group_id": "fresh12:banking77:c851b164c70a9c361826934b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Someone stole my wallet earlier today, not sure exactly when, probably on Piccadilly circus. Can you check if there were any attempts to use the card and obviously block it?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e3701c\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_7ff98d\": \"The user's request concerns atm support.\", \"candidate_6c541e\": \"The user's request concerns disposable card limits.\", \"candidate_c33742\": \"The user's request concerns lost or stolen card.\", \"candidate_f9e95f\": \"The user's request concerns terminate account.\", \"candidate_aab3f9\": \"The user's request concerns getting virtual card.\", \"candidate_c04ede\": \"The user's request concerns failed transfer.\", \"candidate_a4d1b4\": \"The user's request concerns exchange charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e3701c\": 0.0, \"candidate_7ff98d\": 0.0, \"candidate_6c541e\": 0.0, \"candidate_c33742\": 1.0, \"candidate_f9e95f\": 0.0, \"candidate_aab3f9\": 0.0, \"candidate_c04ede\": 0.0, \"candidate_a4d1b4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:bccf66a9bd595be5535ea385:choice", "group_id": "fresh12:banking77:bccf66a9bd595be5535ea385", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How old do you have to be to get an account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ae11b5\": \"The user's request concerns pending top up.\", \"candidate_d6f5f1\": \"The user's request concerns top up by card charge.\", \"candidate_e8babf\": \"The user's request concerns terminate account.\", \"candidate_326b33\": \"The user's request concerns age limit.\", \"candidate_b0368d\": \"The user's request concerns extra charge on statement.\", \"candidate_2f8d9e\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_ba29b0\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_742ffe\": \"The user's request concerns exchange via app.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ae11b5\": 0.0, \"candidate_d6f5f1\": 0.0, \"candidate_e8babf\": 0.0, \"candidate_326b33\": 1.0, \"candidate_b0368d\": 0.0, \"candidate_2f8d9e\": 0.0, \"candidate_ba29b0\": 0.0, \"candidate_742ffe\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c9b0db751004ae4294fbe468:choice", "group_id": "fresh12:banking77:c9b0db751004ae4294fbe468", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How old do we have to be?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dcf131\": \"The user's request concerns supported cards and currencies.\", \"candidate_c0a565\": \"The user's request concerns card linking.\", \"candidate_08ad71\": \"The user's request concerns card payment not recognised.\", \"candidate_dc1b95\": \"The user's request concerns top up by cash or cheque.\", \"candidate_e361a7\": \"The user's request concerns exchange charge.\", \"candidate_4131d9\": \"The user's request concerns atm support.\", \"candidate_8df425\": \"The user's request concerns age limit.\", \"candidate_a24b79\": \"The user's request concerns card delivery estimate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dcf131\": 0.0, \"candidate_c0a565\": 0.0, \"candidate_08ad71\": 0.0, \"candidate_dc1b95\": 0.0, \"candidate_e361a7\": 0.0, \"candidate_4131d9\": 0.0, \"candidate_8df425\": 1.0, \"candidate_a24b79\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3a57fe6088294ad1dbc85dc9:choice", "group_id": "fresh12:banking77:3a57fe6088294ad1dbc85dc9", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How old do my children need to be to open an account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7c0f76\": \"The user's request concerns receiving money.\", \"candidate_501095\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_191386\": \"The user's request concerns why verify identity.\", \"candidate_fd3c35\": \"The user's request concerns age limit.\", \"candidate_4b44bf\": \"The user's request concerns top up reverted.\", \"candidate_4aab60\": \"The user's request concerns cancel transfer.\", \"candidate_a3e60d\": \"The user's request concerns failed transfer.\", \"candidate_b11604\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7c0f76\": 0.0, \"candidate_501095\": 0.0, \"candidate_191386\": 0.0, \"candidate_fd3c35\": 1.0, \"candidate_4b44bf\": 0.0, \"candidate_4aab60\": 0.0, \"candidate_a3e60d\": 0.0, \"candidate_b11604\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:315947fe4fde39e4760283fe:choice", "group_id": "fresh12:banking77:315947fe4fde39e4760283fe", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can my daughter open an account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2aeed4\": \"The user's request concerns get physical card.\", \"candidate_d60419\": \"The user's request concerns supported cards and currencies.\", \"candidate_1b71ec\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_9397ef\": \"The user's request concerns terminate account.\", \"candidate_3b9cc2\": \"The user's request concerns receiving money.\", \"candidate_8d7b93\": \"The user's request concerns apple pay or google pay.\", \"candidate_dee83b\": \"The user's request concerns age limit.\", \"candidate_0f13ec\": \"The user's request concerns failed transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2aeed4\": 0.0, \"candidate_d60419\": 0.0, \"candidate_1b71ec\": 0.0, \"candidate_9397ef\": 0.0, \"candidate_3b9cc2\": 0.0, \"candidate_8d7b93\": 0.0, \"candidate_dee83b\": 1.0, \"candidate_0f13ec\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1637e095aed22f11e3cc43b0:choice", "group_id": "fresh12:banking77:1637e095aed22f11e3cc43b0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How old do I have to be?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_aa5469\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_2a3b26\": \"The user's request concerns pin blocked.\", \"candidate_ea388e\": \"The user's request concerns card swallowed.\", \"candidate_37fe12\": \"The user's request concerns edit personal details.\", \"candidate_7c9b57\": \"The user's request concerns top up by card charge.\", \"candidate_84e2af\": \"The user's request concerns declined cash withdrawal.\", \"candidate_f69998\": \"The user's request concerns atm support.\", \"candidate_d7a9f5\": \"The user's request concerns age limit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_aa5469\": 0.0, \"candidate_2a3b26\": 0.0, \"candidate_ea388e\": 0.0, \"candidate_37fe12\": 0.0, \"candidate_7c9b57\": 0.0, \"candidate_84e2af\": 0.0, \"candidate_f69998\": 0.0, \"candidate_d7a9f5\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:99c6c8c116c444d2cecdb565:choice", "group_id": "fresh12:banking77:99c6c8c116c444d2cecdb565", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What age do my kids need to be to use your service?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_34806d\": \"The user's request concerns top up by cash or cheque.\", \"candidate_857161\": \"The user's request concerns fiat currency support.\", \"candidate_387b8c\": \"The user's request concerns beneficiary not allowed.\", \"candidate_8b89e2\": \"The user's request concerns pending cash withdrawal.\", \"candidate_3baf11\": \"The user's request concerns card payment fee charged.\", \"candidate_5ea4d1\": \"The user's request concerns visa or mastercard.\", \"candidate_e06b6e\": \"The user's request concerns change pin.\", \"candidate_f76da7\": \"The user's request concerns age limit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_34806d\": 0.0, \"candidate_857161\": 0.0, \"candidate_387b8c\": 0.0, \"candidate_8b89e2\": 0.0, \"candidate_3baf11\": 0.0, \"candidate_5ea4d1\": 0.0, \"candidate_e06b6e\": 0.0, \"candidate_f76da7\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:25fd53f80a6471a1eca7bb17:choice", "group_id": "fresh12:banking77:25fd53f80a6471a1eca7bb17", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My account is blocked, how do I log in now\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b588ce\": \"The user's request concerns pin blocked.\", \"candidate_b3b35a\": \"The user's request concerns Refund not showing up.\", \"candidate_4a9fec\": \"The user's request concerns transfer timing.\", \"candidate_1d74fe\": \"The user's request concerns contactless not working.\", \"candidate_2535a7\": \"The user's request concerns supported cards and currencies.\", \"candidate_be84f0\": \"The user's request concerns declined cash withdrawal.\", \"candidate_aabb1d\": \"The user's request concerns card linking.\", \"candidate_99eb22\": \"The user's request concerns lost or stolen card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b588ce\": 1.0, \"candidate_b3b35a\": 0.0, \"candidate_4a9fec\": 0.0, \"candidate_1d74fe\": 0.0, \"candidate_2535a7\": 0.0, \"candidate_be84f0\": 0.0, \"candidate_aabb1d\": 0.0, \"candidate_99eb22\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2506a95dddb8152939911c9f:choice", "group_id": "fresh12:banking77:2506a95dddb8152939911c9f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can I unlock the pin?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_67120f\": \"The user's request concerns card payment fee charged.\", \"candidate_caad99\": \"The user's request concerns exchange via app.\", \"candidate_914f00\": \"The user's request concerns terminate account.\", \"candidate_a66fb4\": \"The user's request concerns get disposable virtual card.\", \"candidate_4b54cf\": \"The user's request concerns disposable card limits.\", \"candidate_8a412d\": \"The user's request concerns card acceptance.\", \"candidate_243b3b\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_b8e343\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_67120f\": 0.0, \"candidate_caad99\": 0.0, \"candidate_914f00\": 0.0, \"candidate_a66fb4\": 0.0, \"candidate_4b54cf\": 0.0, \"candidate_8a412d\": 0.0, \"candidate_243b3b\": 0.0, \"candidate_b8e343\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3b095c8afea9f6597fad818b:choice", "group_id": "fresh12:banking77:3b095c8afea9f6597fad818b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I used the wrong ping too many times and now the account is blocked.  How do I unblock?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b65ba2\": \"The user's request concerns verify top up.\", \"candidate_5b6e0d\": \"The user's request concerns visa or mastercard.\", \"candidate_4d1279\": \"The user's request concerns top up limits.\", \"candidate_74ceab\": \"The user's request concerns country support.\", \"candidate_d8290c\": \"The user's request concerns transfer fee charged.\", \"candidate_c8f600\": \"The user's request concerns pin blocked.\", \"candidate_7b9986\": \"The user's request concerns getting spare card.\", \"candidate_a968a2\": \"The user's request concerns fiat currency support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b65ba2\": 0.0, \"candidate_5b6e0d\": 0.0, \"candidate_4d1279\": 0.0, \"candidate_74ceab\": 0.0, \"candidate_d8290c\": 0.0, \"candidate_c8f600\": 1.0, \"candidate_7b9986\": 0.0, \"candidate_a968a2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ae8037c0169b16421ae109c0:choice", "group_id": "fresh12:banking77:ae8037c0169b16421ae109c0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card's frozen, what can I do?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_96b689\": \"The user's request concerns pin blocked.\", \"candidate_b922e8\": \"The user's request concerns verify my identity.\", \"candidate_f577c9\": \"The user's request concerns automatic top up.\", \"candidate_9b5de6\": \"The user's request concerns transfer into account.\", \"candidate_461b27\": \"The user's request concerns transaction charged twice.\", \"candidate_28e232\": \"The user's request concerns verify top up.\", \"candidate_70f80f\": \"The user's request concerns card swallowed.\", \"candidate_b21e57\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_96b689\": 1.0, \"candidate_b922e8\": 0.0, \"candidate_f577c9\": 0.0, \"candidate_9b5de6\": 0.0, \"candidate_461b27\": 0.0, \"candidate_28e232\": 0.0, \"candidate_70f80f\": 0.0, \"candidate_b21e57\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:128ff675dca36dbdda767634:choice", "group_id": "fresh12:banking77:128ff675dca36dbdda767634", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can you unlock my pin? I think I entered the wrong pin too many times.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f475a3\": \"The user's request concerns activate my card.\", \"candidate_2c332e\": \"The user's request concerns getting virtual card.\", \"candidate_14cd4b\": \"The user's request concerns cash withdrawal charge.\", \"candidate_260ca0\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_a8151e\": \"The user's request concerns get disposable virtual card.\", \"candidate_c636b6\": \"The user's request concerns card about to expire.\", \"candidate_3091a9\": \"The user's request concerns pin blocked.\", \"candidate_23acf1\": \"The user's request concerns transfer fee charged.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f475a3\": 0.0, \"candidate_2c332e\": 0.0, \"candidate_14cd4b\": 0.0, \"candidate_260ca0\": 0.0, \"candidate_a8151e\": 0.0, \"candidate_c636b6\": 0.0, \"candidate_3091a9\": 1.0, \"candidate_23acf1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c25441dae3784a02d6208d9f:choice", "group_id": "fresh12:banking77:c25441dae3784a02d6208d9f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What should I do if I've tried to enter my PIN too often?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c4f1c4\": \"The user's request concerns virtual card not working.\", \"candidate_a1aac8\": \"The user's request concerns card not working.\", \"candidate_2f89ee\": \"The user's request concerns card arrival.\", \"candidate_a352e7\": \"The user's request concerns lost or stolen phone.\", \"candidate_e59428\": \"The user's request concerns age limit.\", \"candidate_9d53d0\": \"The user's request concerns transfer fee charged.\", \"candidate_1fb0cc\": \"The user's request concerns failed transfer.\", \"candidate_5f5cd1\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c4f1c4\": 0.0, \"candidate_a1aac8\": 0.0, \"candidate_2f89ee\": 0.0, \"candidate_a352e7\": 0.0, \"candidate_e59428\": 0.0, \"candidate_9d53d0\": 0.0, \"candidate_1fb0cc\": 0.0, \"candidate_5f5cd1\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:bf4112f0019d4e1b5cd6c79e:choice", "group_id": "fresh12:banking77:bf4112f0019d4e1b5cd6c79e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I wanted to use my contactless at the metro today but it didn't accept it!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_661478\": \"The user's request concerns pending card payment.\", \"candidate_c3aa3b\": \"The user's request concerns verify source of funds.\", \"candidate_d29eeb\": \"The user's request concerns Refund not showing up.\", \"candidate_f51223\": \"The user's request concerns exchange charge.\", \"candidate_6d8793\": \"The user's request concerns contactless not working.\", \"candidate_d3f739\": \"The user's request concerns pending top up.\", \"candidate_44f24d\": \"The user's request concerns card linking.\", \"candidate_79f30d\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_661478\": 0.0, \"candidate_c3aa3b\": 0.0, \"candidate_d29eeb\": 0.0, \"candidate_f51223\": 0.0, \"candidate_6d8793\": 1.0, \"candidate_d3f739\": 0.0, \"candidate_44f24d\": 0.0, \"candidate_79f30d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d5985c3e51254bbede8e82fd:choice", "group_id": "fresh12:banking77:d5985c3e51254bbede8e82fd", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I wanted to use my card contactless but it doesn't seem to be working, what could be the problem?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b301ea\": \"The user's request concerns top up by cash or cheque.\", \"candidate_00bc60\": \"The user's request concerns card payment not recognised.\", \"candidate_ab1b70\": \"The user's request concerns contactless not working.\", \"candidate_d86384\": \"The user's request concerns card payment fee charged.\", \"candidate_7f2edf\": \"The user's request concerns transfer timing.\", \"candidate_6544ad\": \"The user's request concerns Refund not showing up.\", \"candidate_e3e0b3\": \"The user's request concerns automatic top up.\", \"candidate_e55c57\": \"The user's request concerns wrong amount of cash received.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b301ea\": 0.0, \"candidate_00bc60\": 0.0, \"candidate_ab1b70\": 1.0, \"candidate_d86384\": 0.0, \"candidate_7f2edf\": 0.0, \"candidate_6544ad\": 0.0, \"candidate_e3e0b3\": 0.0, \"candidate_e55c57\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1370fb4022441898dd2d354a:choice", "group_id": "fresh12:banking77:1370fb4022441898dd2d354a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Should i reinstall the payment app?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_57a3a4\": \"The user's request concerns exchange rate.\", \"candidate_bd3e42\": \"The user's request concerns get physical card.\", \"candidate_c61df7\": \"The user's request concerns terminate account.\", \"candidate_bd2669\": \"The user's request concerns card acceptance.\", \"candidate_6c6d79\": \"The user's request concerns exchange charge.\", \"candidate_04cdb7\": \"The user's request concerns top up by card charge.\", \"candidate_bb5167\": \"The user's request concerns verify my identity.\", \"candidate_4ba2f0\": \"The user's request concerns contactless not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_57a3a4\": 0.0, \"candidate_bd3e42\": 0.0, \"candidate_c61df7\": 0.0, \"candidate_bd2669\": 0.0, \"candidate_6c6d79\": 0.0, \"candidate_04cdb7\": 0.0, \"candidate_bb5167\": 0.0, \"candidate_4ba2f0\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9d13979174daa5191cfe9c04:choice", "group_id": "fresh12:banking77:9d13979174daa5191cfe9c04", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can I make my contactless work for the metro?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_efa224\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_3cd42d\": \"The user's request concerns contactless not working.\", \"candidate_af0f3f\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_ba9b25\": \"The user's request concerns passcode forgotten.\", \"candidate_fd69ee\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_f572a1\": \"The user's request concerns activate my card.\", \"candidate_54c335\": \"The user's request concerns card arrival.\", \"candidate_e58845\": \"The user's request concerns balance not updated after bank transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_efa224\": 0.0, \"candidate_3cd42d\": 1.0, \"candidate_af0f3f\": 0.0, \"candidate_ba9b25\": 0.0, \"candidate_fd69ee\": 0.0, \"candidate_f572a1\": 0.0, \"candidate_54c335\": 0.0, \"candidate_e58845\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1cc1bce474ec03b00678c570:choice", "group_id": "fresh12:banking77:1cc1bce474ec03b00678c570", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"what is the charges for new card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b61006\": \"The user's request concerns card payment not recognised.\", \"candidate_3f1c5f\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_23e774\": \"The user's request concerns contactless not working.\", \"candidate_a9588e\": \"The user's request concerns getting virtual card.\", \"candidate_d3dbca\": \"The user's request concerns card about to expire.\", \"candidate_98ec38\": \"The user's request concerns unable to verify identity.\", \"candidate_2dca59\": \"The user's request concerns transfer into account.\", \"candidate_1c795a\": \"The user's request concerns top up limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b61006\": 0.0, \"candidate_3f1c5f\": 0.0, \"candidate_23e774\": 1.0, \"candidate_a9588e\": 0.0, \"candidate_d3dbca\": 0.0, \"candidate_98ec38\": 0.0, \"candidate_2dca59\": 0.0, \"candidate_1c795a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:786b2d9adce8c5c29c032795:choice", "group_id": "fresh12:banking77:786b2d9adce8c5c29c032795", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why wouldn't the contactless payment work when I tried to pay at the bus today?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8ba76b\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_b3e7be\": \"The user's request concerns contactless not working.\", \"candidate_bd4461\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_a7b0da\": \"The user's request concerns top up by card charge.\", \"candidate_1fc15a\": \"The user's request concerns get disposable virtual card.\", \"candidate_e58bb7\": \"The user's request concerns getting spare card.\", \"candidate_7c7d3e\": \"The user's request concerns card payment not recognised.\", \"candidate_cf56bf\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8ba76b\": 0.0, \"candidate_b3e7be\": 1.0, \"candidate_bd4461\": 0.0, \"candidate_a7b0da\": 0.0, \"candidate_1fc15a\": 0.0, \"candidate_e58bb7\": 0.0, \"candidate_7c7d3e\": 0.0, \"candidate_cf56bf\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4186924043d816d787e5546e:choice", "group_id": "fresh12:banking77:4186924043d816d787e5546e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Please tell me about SWIFT transfers at this bank.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c62215\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_83a877\": \"The user's request concerns automatic top up.\", \"candidate_d51266\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_6858fd\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_7fcfc1\": \"The user's request concerns transaction charged twice.\", \"candidate_8e1e21\": \"The user's request concerns top up failed.\", \"candidate_83a60f\": \"The user's request concerns beneficiary not allowed.\", \"candidate_d768d7\": \"The user's request concerns pending transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c62215\": 0.0, \"candidate_83a877\": 0.0, \"candidate_d51266\": 0.0, \"candidate_6858fd\": 1.0, \"candidate_7fcfc1\": 0.0, \"candidate_8e1e21\": 0.0, \"candidate_83a60f\": 0.0, \"candidate_d768d7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:42aec171570d7e0001b9a8ab:choice", "group_id": "fresh12:banking77:42aec171570d7e0001b9a8ab", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Will you deal with SWIFT transfers?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7b6331\": \"The user's request concerns card about to expire.\", \"candidate_40480c\": \"The user's request concerns request refund.\", \"candidate_3e499a\": \"The user's request concerns disposable card limits.\", \"candidate_07399d\": \"The user's request concerns Refund not showing up.\", \"candidate_422146\": \"The user's request concerns top up reverted.\", \"candidate_96aa9a\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_04c494\": \"The user's request concerns contactless not working.\", \"candidate_833645\": \"The user's request concerns lost or stolen card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7b6331\": 0.0, \"candidate_40480c\": 0.0, \"candidate_3e499a\": 0.0, \"candidate_07399d\": 0.0, \"candidate_422146\": 0.0, \"candidate_96aa9a\": 1.0, \"candidate_04c494\": 0.0, \"candidate_833645\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0ba94c0a58502309825505a1:choice", "group_id": "fresh12:banking77:0ba94c0a58502309825505a1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I also transfer with SWIFT?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_81dc39\": \"The user's request concerns top up failed.\", \"candidate_f47ac3\": \"The user's request concerns order physical card.\", \"candidate_1ac7ce\": \"The user's request concerns transfer timing.\", \"candidate_4dea79\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_c9e836\": \"The user's request concerns verify source of funds.\", \"candidate_8ea44d\": \"The user's request concerns contactless not working.\", \"candidate_e8392c\": \"The user's request concerns getting spare card.\", \"candidate_7f89ef\": \"The user's request concerns verify my identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_81dc39\": 0.0, \"candidate_f47ac3\": 0.0, \"candidate_1ac7ce\": 0.0, \"candidate_4dea79\": 1.0, \"candidate_c9e836\": 0.0, \"candidate_8ea44d\": 0.0, \"candidate_e8392c\": 0.0, \"candidate_7f89ef\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5d7075e43255edcbb652dce4:choice", "group_id": "fresh12:banking77:5d7075e43255edcbb652dce4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Will I be charged for a SEPA transfer?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5764fd\": \"The user's request concerns compromised card.\", \"candidate_5ba773\": \"The user's request concerns transfer fee charged.\", \"candidate_e331ef\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_b78c8f\": \"The user's request concerns transfer timing.\", \"candidate_1b096e\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_8cfd1e\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_b93ee8\": \"The user's request concerns failed transfer.\", \"candidate_8f3b57\": \"The user's request concerns card payment fee charged.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5764fd\": 0.0, \"candidate_5ba773\": 0.0, \"candidate_e331ef\": 1.0, \"candidate_b78c8f\": 0.0, \"candidate_1b096e\": 0.0, \"candidate_8cfd1e\": 0.0, \"candidate_b93ee8\": 0.0, \"candidate_8f3b57\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6d8eb35d1a6678635933c231:choice", "group_id": "fresh12:banking77:6d8eb35d1a6678635933c231", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What are the fees for top-ups?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_af4e52\": \"The user's request concerns lost or stolen card.\", \"candidate_a61b58\": \"The user's request concerns cash withdrawal charge.\", \"candidate_e947f3\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_c17c4a\": \"The user's request concerns exchange rate.\", \"candidate_998927\": \"The user's request concerns card delivery estimate.\", \"candidate_e1917a\": \"The user's request concerns pending cash withdrawal.\", \"candidate_c104aa\": \"The user's request concerns getting virtual card.\", \"candidate_5c6457\": \"The user's request concerns order physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_af4e52\": 0.0, \"candidate_a61b58\": 0.0, \"candidate_e947f3\": 1.0, \"candidate_c17c4a\": 0.0, \"candidate_998927\": 0.0, \"candidate_e1917a\": 0.0, \"candidate_c104aa\": 0.0, \"candidate_5c6457\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1f830de8b8bd65d57c8c1c3b:choice", "group_id": "fresh12:banking77:1f830de8b8bd65d57c8c1c3b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I would like to refill my account using SWIFT.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_18737f\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_5c4822\": \"The user's request concerns top up reverted.\", \"candidate_87b0d8\": \"The user's request concerns cash withdrawal charge.\", \"candidate_c694c2\": \"The user's request concerns declined card payment.\", \"candidate_ecfb6b\": \"The user's request concerns card delivery estimate.\", \"candidate_7c5b6a\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_d132f9\": \"The user's request concerns why verify identity.\", \"candidate_f715e2\": \"The user's request concerns wrong exchange rate for cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_18737f\": 0.0, \"candidate_5c4822\": 0.0, \"candidate_87b0d8\": 0.0, \"candidate_c694c2\": 0.0, \"candidate_ecfb6b\": 0.0, \"candidate_7c5b6a\": 1.0, \"candidate_d132f9\": 0.0, \"candidate_f715e2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:656b7835b73218facedf79d0:choice", "group_id": "fresh12:banking77:656b7835b73218facedf79d0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What would lead to my top up still pending?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c3c4f8\": \"The user's request concerns visa or mastercard.\", \"candidate_28cc91\": \"The user's request concerns card payment not recognised.\", \"candidate_a91f9f\": \"The user's request concerns card about to expire.\", \"candidate_bdbac3\": \"The user's request concerns declined transfer.\", \"candidate_8d758b\": \"The user's request concerns apple pay or google pay.\", \"candidate_a486f9\": \"The user's request concerns why verify identity.\", \"candidate_4c7351\": \"The user's request concerns verify source of funds.\", \"candidate_a94a72\": \"The user's request concerns pending top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c3c4f8\": 0.0, \"candidate_28cc91\": 0.0, \"candidate_a91f9f\": 0.0, \"candidate_bdbac3\": 0.0, \"candidate_8d758b\": 0.0, \"candidate_a486f9\": 0.0, \"candidate_4c7351\": 0.0, \"candidate_a94a72\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8c4be00ea3f6d62fb5edf5e8:choice", "group_id": "fresh12:banking77:8c4be00ea3f6d62fb5edf5e8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The top-up has a definite problem! I regularly use your service and haven't had a problem before, but now the top-up just shows pending. Please tell my why it continues to show pending.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_042020\": \"The user's request concerns declined cash withdrawal.\", \"candidate_0d8c31\": \"The user's request concerns atm support.\", \"candidate_31dffb\": \"The user's request concerns top up reverted.\", \"candidate_ac6d9d\": \"The user's request concerns visa or mastercard.\", \"candidate_457945\": \"The user's request concerns transaction charged twice.\", \"candidate_6c350c\": \"The user's request concerns pending top up.\", \"candidate_d50bb0\": \"The user's request concerns card arrival.\", \"candidate_eec639\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_042020\": 0.0, \"candidate_0d8c31\": 0.0, \"candidate_31dffb\": 0.0, \"candidate_ac6d9d\": 0.0, \"candidate_457945\": 0.0, \"candidate_6c350c\": 1.0, \"candidate_d50bb0\": 0.0, \"candidate_eec639\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:72a4167c25e65d681fdcddb1:choice", "group_id": "fresh12:banking77:72a4167c25e65d681fdcddb1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can you look into my top up please.  I made it over three hours ago and yet it's still pending.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e788ee\": \"The user's request concerns pending top up.\", \"candidate_fca3cb\": \"The user's request concerns disposable card limits.\", \"candidate_9e905e\": \"The user's request concerns compromised card.\", \"candidate_9eb9e4\": \"The user's request concerns cash withdrawal charge.\", \"candidate_d2e2e0\": \"The user's request concerns pending card payment.\", \"candidate_4d785a\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_7fb083\": \"The user's request concerns getting virtual card.\", \"candidate_ba8f8b\": \"The user's request concerns balance not updated after bank transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e788ee\": 1.0, \"candidate_fca3cb\": 0.0, \"candidate_9e905e\": 0.0, \"candidate_9eb9e4\": 0.0, \"candidate_d2e2e0\": 0.0, \"candidate_4d785a\": 0.0, \"candidate_7fb083\": 0.0, \"candidate_ba8f8b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5e21178d1e156f21903c620f:choice", "group_id": "fresh12:banking77:5e21178d1e156f21903c620f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card was topped this morning but I can't see the funds. Why didn't it complete?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2b083f\": \"The user's request concerns visa or mastercard.\", \"candidate_c2c53d\": \"The user's request concerns disposable card limits.\", \"candidate_5ca9b5\": \"The user's request concerns pending transfer.\", \"candidate_42a75e\": \"The user's request concerns request refund.\", \"candidate_22381b\": \"The user's request concerns pending top up.\", \"candidate_3783b2\": \"The user's request concerns pending card payment.\", \"candidate_59b2cf\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_8bff1f\": \"The user's request concerns order physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2b083f\": 0.0, \"candidate_c2c53d\": 0.0, \"candidate_5ca9b5\": 0.0, \"candidate_42a75e\": 0.0, \"candidate_22381b\": 1.0, \"candidate_3783b2\": 0.0, \"candidate_59b2cf\": 0.0, \"candidate_8bff1f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c3592e229e4cf5cd788fa65a:choice", "group_id": "fresh12:banking77:c3592e229e4cf5cd788fa65a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why hasn't my top-up been completed?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_355cc1\": \"The user's request concerns beneficiary not allowed.\", \"candidate_e63ba8\": \"The user's request concerns pending top up.\", \"candidate_462cab\": \"The user's request concerns request refund.\", \"candidate_bcccb8\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_452959\": \"The user's request concerns card arrival.\", \"candidate_7d71c1\": \"The user's request concerns Refund not showing up.\", \"candidate_ab62fb\": \"The user's request concerns top up reverted.\", \"candidate_c32c26\": \"The user's request concerns declined transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_355cc1\": 0.0, \"candidate_e63ba8\": 1.0, \"candidate_462cab\": 0.0, \"candidate_bcccb8\": 0.0, \"candidate_452959\": 0.0, \"candidate_7d71c1\": 0.0, \"candidate_ab62fb\": 0.0, \"candidate_c32c26\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:60eded18d5b39452dd8c2fbb:choice", "group_id": "fresh12:banking77:60eded18d5b39452dd8c2fbb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there something wrong with your website? I tried topping up my account and it's been close to two hours now and it's still at \\\"pending\\\" for some reason. I just joined with you guys and this is the first attempt at this, so maybe I'm wrong, but shouldn't this be instant?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_629df0\": \"The user's request concerns pending cash withdrawal.\", \"candidate_5d90e2\": \"The user's request concerns pending top up.\", \"candidate_a1a7d3\": \"The user's request concerns card acceptance.\", \"candidate_b772bb\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_e9cbe0\": \"The user's request concerns lost or stolen phone.\", \"candidate_35fd21\": \"The user's request concerns exchange charge.\", \"candidate_750fd4\": \"The user's request concerns cancel transfer.\", \"candidate_b2d4b7\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_629df0\": 0.0, \"candidate_5d90e2\": 1.0, \"candidate_a1a7d3\": 0.0, \"candidate_b772bb\": 0.0, \"candidate_e9cbe0\": 0.0, \"candidate_35fd21\": 0.0, \"candidate_750fd4\": 0.0, \"candidate_b2d4b7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5f3737a23006d301ade67858:choice", "group_id": "fresh12:banking77:5f3737a23006d301ade67858", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I want to cancel a transfer right away! Can you help?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_08617b\": \"The user's request concerns getting spare card.\", \"candidate_b5c259\": \"The user's request concerns transfer fee charged.\", \"candidate_7ea1a9\": \"The user's request concerns passcode forgotten.\", \"candidate_e2a3fa\": \"The user's request concerns wrong amount of cash received.\", \"candidate_ff78f3\": \"The user's request concerns receiving money.\", \"candidate_f752ef\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_249110\": \"The user's request concerns cancel transfer.\", \"candidate_75ff0c\": \"The user's request concerns top up failed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_08617b\": 0.0, \"candidate_b5c259\": 0.0, \"candidate_7ea1a9\": 0.0, \"candidate_e2a3fa\": 0.0, \"candidate_ff78f3\": 0.0, \"candidate_f752ef\": 0.0, \"candidate_249110\": 1.0, \"candidate_75ff0c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0262a0d3c44d33a1ff529499:choice", "group_id": "fresh12:banking77:0262a0d3c44d33a1ff529499", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need to cancel my recent transfer immediately, I made a mistake there, please help quickly before it goes through\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_69d381\": \"The user's request concerns country support.\", \"candidate_9e7222\": \"The user's request concerns top up reverted.\", \"candidate_d205cc\": \"The user's request concerns transfer fee charged.\", \"candidate_0454f8\": \"The user's request concerns lost or stolen card.\", \"candidate_461fac\": \"The user's request concerns disposable card limits.\", \"candidate_64706e\": \"The user's request concerns pending top up.\", \"candidate_d6406a\": \"The user's request concerns passcode forgotten.\", \"candidate_189ef5\": \"The user's request concerns cancel transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_69d381\": 0.0, \"candidate_9e7222\": 0.0, \"candidate_d205cc\": 0.0, \"candidate_0454f8\": 0.0, \"candidate_461fac\": 0.0, \"candidate_64706e\": 0.0, \"candidate_d6406a\": 0.0, \"candidate_189ef5\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ad36e6f8de3fd6c91edd8462:choice", "group_id": "fresh12:banking77:ad36e6f8de3fd6c91edd8462", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I want to go back on what I did this morning\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1a6298\": \"The user's request concerns topping up by card.\", \"candidate_98ba85\": \"The user's request concerns lost or stolen card.\", \"candidate_cdb495\": \"The user's request concerns failed transfer.\", \"candidate_6cf6ba\": \"The user's request concerns cancel transfer.\", \"candidate_baeab0\": \"The user's request concerns disposable card limits.\", \"candidate_e63059\": \"The user's request concerns card acceptance.\", \"candidate_b16486\": \"The user's request concerns declined cash withdrawal.\", \"candidate_255648\": \"The user's request concerns top up limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1a6298\": 0.0, \"candidate_98ba85\": 0.0, \"candidate_cdb495\": 0.0, \"candidate_6cf6ba\": 1.0, \"candidate_baeab0\": 0.0, \"candidate_e63059\": 0.0, \"candidate_b16486\": 0.0, \"candidate_255648\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0959ef9e98b60312cec675ca:choice", "group_id": "fresh12:banking77:0959ef9e98b60312cec675ca", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I made a mistake and need to cancel a transaction\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_04efe4\": \"The user's request concerns lost or stolen card.\", \"candidate_83fc6c\": \"The user's request concerns top up failed.\", \"candidate_f21d09\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_a2825e\": \"The user's request concerns lost or stolen phone.\", \"candidate_bef75e\": \"The user's request concerns country support.\", \"candidate_43c256\": \"The user's request concerns transfer not received by recipient.\", \"candidate_75d59e\": \"The user's request concerns cancel transfer.\", \"candidate_089a86\": \"The user's request concerns declined cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_04efe4\": 0.0, \"candidate_83fc6c\": 0.0, \"candidate_f21d09\": 0.0, \"candidate_a2825e\": 0.0, \"candidate_bef75e\": 0.0, \"candidate_43c256\": 0.0, \"candidate_75d59e\": 1.0, \"candidate_089a86\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1a037e82eeaa9c358053ef8f:choice", "group_id": "fresh12:banking77:1a037e82eeaa9c358053ef8f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I cancel a transfer\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_74a3c8\": \"The user's request concerns receiving money.\", \"candidate_3ba4d7\": \"The user's request concerns pending card payment.\", \"candidate_c98142\": \"The user's request concerns card payment not recognised.\", \"candidate_dbde98\": \"The user's request concerns exchange rate.\", \"candidate_d5ccf2\": \"The user's request concerns cash withdrawal charge.\", \"candidate_a587b8\": \"The user's request concerns cancel transfer.\", \"candidate_fd780b\": \"The user's request concerns transfer fee charged.\", \"candidate_c24fd9\": \"The user's request concerns compromised card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_74a3c8\": 0.0, \"candidate_3ba4d7\": 0.0, \"candidate_c98142\": 0.0, \"candidate_dbde98\": 0.0, \"candidate_d5ccf2\": 0.0, \"candidate_a587b8\": 1.0, \"candidate_fd780b\": 0.0, \"candidate_c24fd9\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0a7d3b81b923499ccadb65d7:choice", "group_id": "fresh12:banking77:0a7d3b81b923499ccadb65d7", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I transferred the wrong amount and would like to cancel the transaction.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e98195\": \"The user's request concerns top up reverted.\", \"candidate_dacd45\": \"The user's request concerns verify top up.\", \"candidate_225a0e\": \"The user's request concerns card swallowed.\", \"candidate_d8b7f0\": \"The user's request concerns card payment not recognised.\", \"candidate_b88b33\": \"The user's request concerns beneficiary not allowed.\", \"candidate_21b84d\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_9f3965\": \"The user's request concerns automatic top up.\", \"candidate_535592\": \"The user's request concerns cancel transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e98195\": 0.0, \"candidate_dacd45\": 0.0, \"candidate_225a0e\": 0.0, \"candidate_d8b7f0\": 0.0, \"candidate_b88b33\": 0.0, \"candidate_21b84d\": 0.0, \"candidate_9f3965\": 0.0, \"candidate_535592\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:aef4e9d8bae85260d4ddd7ea:choice", "group_id": "fresh12:banking77:aef4e9d8bae85260d4ddd7ea", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a limit on top-ups?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_217ed4\": \"The user's request concerns apple pay or google pay.\", \"candidate_73dd04\": \"The user's request concerns get physical card.\", \"candidate_33b252\": \"The user's request concerns top up limits.\", \"candidate_93feb5\": \"The user's request concerns getting virtual card.\", \"candidate_3fdbd8\": \"The user's request concerns card about to expire.\", \"candidate_fba0c9\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_de0ea7\": \"The user's request concerns pending top up.\", \"candidate_52ed98\": \"The user's request concerns Refund not showing up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_217ed4\": 0.0, \"candidate_73dd04\": 0.0, \"candidate_33b252\": 1.0, \"candidate_93feb5\": 0.0, \"candidate_3fdbd8\": 0.0, \"candidate_fba0c9\": 0.0, \"candidate_de0ea7\": 0.0, \"candidate_52ed98\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4be949410a1022dcaa4c6345:choice", "group_id": "fresh12:banking77:4be949410a1022dcaa4c6345", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What is the top-ups limit?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0f282b\": \"The user's request concerns top up limits.\", \"candidate_546c20\": \"The user's request concerns request refund.\", \"candidate_34c89c\": \"The user's request concerns disposable card limits.\", \"candidate_b98db2\": \"The user's request concerns getting virtual card.\", \"candidate_50e07c\": \"The user's request concerns verify source of funds.\", \"candidate_d33e46\": \"The user's request concerns unable to verify identity.\", \"candidate_759e6c\": \"The user's request concerns apple pay or google pay.\", \"candidate_ffe099\": \"The user's request concerns visa or mastercard.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0f282b\": 1.0, \"candidate_546c20\": 0.0, \"candidate_34c89c\": 0.0, \"candidate_b98db2\": 0.0, \"candidate_50e07c\": 0.0, \"candidate_d33e46\": 0.0, \"candidate_759e6c\": 0.0, \"candidate_ffe099\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:36b57f35be2f85294f926cdc:choice", "group_id": "fresh12:banking77:36b57f35be2f85294f926cdc", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can I top-up my limit?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f850e7\": \"The user's request concerns declined card payment.\", \"candidate_7053d6\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_f3d5fb\": \"The user's request concerns top up limits.\", \"candidate_52fc91\": \"The user's request concerns top up failed.\", \"candidate_50f05e\": \"The user's request concerns passcode forgotten.\", \"candidate_91aca4\": \"The user's request concerns declined transfer.\", \"candidate_da8756\": \"The user's request concerns top up by card charge.\", \"candidate_5bd6e1\": \"The user's request concerns request refund.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f850e7\": 0.0, \"candidate_7053d6\": 0.0, \"candidate_f3d5fb\": 1.0, \"candidate_52fc91\": 0.0, \"candidate_50f05e\": 0.0, \"candidate_91aca4\": 0.0, \"candidate_da8756\": 0.0, \"candidate_5bd6e1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:47892843ab3c9206035e63bd:choice", "group_id": "fresh12:banking77:47892843ab3c9206035e63bd", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can you tell me the limits for top ups?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c13cfa\": \"The user's request concerns card payment fee charged.\", \"candidate_f7f456\": \"The user's request concerns reverted card payment?.\", \"candidate_56bebb\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_e005ee\": \"The user's request concerns pending cash withdrawal.\", \"candidate_5bc95c\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_aba376\": \"The user's request concerns wrong amount of cash received.\", \"candidate_15691a\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_8d6974\": \"The user's request concerns top up limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c13cfa\": 0.0, \"candidate_f7f456\": 0.0, \"candidate_56bebb\": 0.0, \"candidate_e005ee\": 0.0, \"candidate_5bc95c\": 0.0, \"candidate_aba376\": 0.0, \"candidate_15691a\": 0.0, \"candidate_8d6974\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ae87d91a4d382e8e67c9071d:choice", "group_id": "fresh12:banking77:ae87d91a4d382e8e67c9071d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Are there limits to top-ups?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5e321f\": \"The user's request concerns reverted card payment?.\", \"candidate_218cec\": \"The user's request concerns wrong amount of cash received.\", \"candidate_7b8ac8\": \"The user's request concerns visa or mastercard.\", \"candidate_2a29fb\": \"The user's request concerns card not working.\", \"candidate_36a163\": \"The user's request concerns top up limits.\", \"candidate_be89ea\": \"The user's request concerns lost or stolen card.\", \"candidate_3d99a7\": \"The user's request concerns pending cash withdrawal.\", \"candidate_1b9075\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5e321f\": 0.0, \"candidate_218cec\": 0.0, \"candidate_7b8ac8\": 0.0, \"candidate_2a29fb\": 0.0, \"candidate_36a163\": 1.0, \"candidate_be89ea\": 0.0, \"candidate_3d99a7\": 0.0, \"candidate_1b9075\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:801d1ebaea1f5de162288dbb:choice", "group_id": "fresh12:banking77:801d1ebaea1f5de162288dbb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a limit for top-ups?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_793753\": \"The user's request concerns exchange charge.\", \"candidate_1e99cb\": \"The user's request concerns age limit.\", \"candidate_91e84c\": \"The user's request concerns Refund not showing up.\", \"candidate_549e47\": \"The user's request concerns automatic top up.\", \"candidate_25e7b3\": \"The user's request concerns top up limits.\", \"candidate_83944a\": \"The user's request concerns verify my identity.\", \"candidate_8c76d8\": \"The user's request concerns card linking.\", \"candidate_c303c5\": \"The user's request concerns country support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_793753\": 0.0, \"candidate_1e99cb\": 0.0, \"candidate_91e84c\": 0.0, \"candidate_549e47\": 0.0, \"candidate_25e7b3\": 1.0, \"candidate_83944a\": 0.0, \"candidate_8c76d8\": 0.0, \"candidate_c303c5\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2d6a47768957739b03092faa:choice", "group_id": "fresh12:banking77:2d6a47768957739b03092faa", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I didn't receive the right amount of cash back\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ccf41f\": \"The user's request concerns pending cash withdrawal.\", \"candidate_357a9f\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_75d30f\": \"The user's request concerns top up by card charge.\", \"candidate_3b60a2\": \"The user's request concerns pending card payment.\", \"candidate_57a94a\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_7bc530\": \"The user's request concerns wrong amount of cash received.\", \"candidate_c5079d\": \"The user's request concerns pending top up.\", \"candidate_b49815\": \"The user's request concerns virtual card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ccf41f\": 0.0, \"candidate_357a9f\": 0.0, \"candidate_75d30f\": 0.0, \"candidate_3b60a2\": 0.0, \"candidate_57a94a\": 0.0, \"candidate_7bc530\": 1.0, \"candidate_c5079d\": 0.0, \"candidate_b49815\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d5f87b07398af90122a53402:choice", "group_id": "fresh12:banking77:d5f87b07398af90122a53402", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I asked for $100 cash but only got $20?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e6c0c7\": \"The user's request concerns wrong amount of cash received.\", \"candidate_2199ef\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_79d5df\": \"The user's request concerns lost or stolen phone.\", \"candidate_9ff4a8\": \"The user's request concerns request refund.\", \"candidate_bd6ff0\": \"The user's request concerns verify my identity.\", \"candidate_b19a63\": \"The user's request concerns compromised card.\", \"candidate_f47f8b\": \"The user's request concerns failed transfer.\", \"candidate_de0fec\": \"The user's request concerns wrong exchange rate for cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e6c0c7\": 1.0, \"candidate_2199ef\": 0.0, \"candidate_79d5df\": 0.0, \"candidate_9ff4a8\": 0.0, \"candidate_bd6ff0\": 0.0, \"candidate_b19a63\": 0.0, \"candidate_f47f8b\": 0.0, \"candidate_de0fec\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:15f6b81b2c92072955255d76:choice", "group_id": "fresh12:banking77:15f6b81b2c92072955255d76", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The ATM gave me the wrong amount of cash today when I went to withdraw. The app says it was more.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3e070f\": \"The user's request concerns why verify identity.\", \"candidate_489cf7\": \"The user's request concerns pending top up.\", \"candidate_38a77f\": \"The user's request concerns beneficiary not allowed.\", \"candidate_f6c02f\": \"The user's request concerns transfer not received by recipient.\", \"candidate_98ee01\": \"The user's request concerns transfer fee charged.\", \"candidate_957b9e\": \"The user's request concerns declined card payment.\", \"candidate_b93be3\": \"The user's request concerns lost or stolen phone.\", \"candidate_e42562\": \"The user's request concerns wrong amount of cash received.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3e070f\": 0.0, \"candidate_489cf7\": 0.0, \"candidate_38a77f\": 0.0, \"candidate_f6c02f\": 0.0, \"candidate_98ee01\": 0.0, \"candidate_957b9e\": 0.0, \"candidate_b93be3\": 0.0, \"candidate_e42562\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:311559c2e7f01b12cd253e4f:choice", "group_id": "fresh12:banking77:311559c2e7f01b12cd253e4f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why did I not get all the cash I asked for?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e1b1c7\": \"The user's request concerns getting spare card.\", \"candidate_fc8ec1\": \"The user's request concerns supported cards and currencies.\", \"candidate_c18c6c\": \"The user's request concerns lost or stolen phone.\", \"candidate_048cbd\": \"The user's request concerns top up failed.\", \"candidate_1b1f22\": \"The user's request concerns terminate account.\", \"candidate_1381a1\": \"The user's request concerns pending card payment.\", \"candidate_2c9a3d\": \"The user's request concerns passcode forgotten.\", \"candidate_90d34a\": \"The user's request concerns wrong amount of cash received.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e1b1c7\": 0.0, \"candidate_fc8ec1\": 0.0, \"candidate_c18c6c\": 0.0, \"candidate_048cbd\": 0.0, \"candidate_1b1f22\": 0.0, \"candidate_1381a1\": 0.0, \"candidate_2c9a3d\": 0.0, \"candidate_90d34a\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:64b10d443626858a6c652c31:choice", "group_id": "fresh12:banking77:64b10d443626858a6c652c31", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Something went wrong with my withdrawal. The amount of money isn't right.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7c5e01\": \"The user's request concerns wrong amount of cash received.\", \"candidate_9fa310\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_68694b\": \"The user's request concerns top up reverted.\", \"candidate_f940c4\": \"The user's request concerns declined card payment.\", \"candidate_4be451\": \"The user's request concerns edit personal details.\", \"candidate_2dff07\": \"The user's request concerns transfer into account.\", \"candidate_460d56\": \"The user's request concerns terminate account.\", \"candidate_79d03b\": \"The user's request concerns reverted card payment?.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7c5e01\": 1.0, \"candidate_9fa310\": 0.0, \"candidate_68694b\": 0.0, \"candidate_f940c4\": 0.0, \"candidate_4be451\": 0.0, \"candidate_2dff07\": 0.0, \"candidate_460d56\": 0.0, \"candidate_79d03b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c76cbd31a32685fb97f8419d:choice", "group_id": "fresh12:banking77:c76cbd31a32685fb97f8419d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The ATM didn't give me the the withdraw amount requested.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6ba587\": \"The user's request concerns passcode forgotten.\", \"candidate_3a1384\": \"The user's request concerns wrong amount of cash received.\", \"candidate_25a150\": \"The user's request concerns disposable card limits.\", \"candidate_3815a2\": \"The user's request concerns card about to expire.\", \"candidate_a9aba8\": \"The user's request concerns pending cash withdrawal.\", \"candidate_9310cb\": \"The user's request concerns get physical card.\", \"candidate_af7c01\": \"The user's request concerns exchange charge.\", \"candidate_a89996\": \"The user's request concerns change pin.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6ba587\": 0.0, \"candidate_3a1384\": 1.0, \"candidate_25a150\": 0.0, \"candidate_3815a2\": 0.0, \"candidate_a9aba8\": 0.0, \"candidate_9310cb\": 0.0, \"candidate_af7c01\": 0.0, \"candidate_a89996\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2707fbc98a8f15d3103386de:choice", "group_id": "fresh12:banking77:2707fbc98a8f15d3103386de", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why do I get charged additional fees on some payments, but not others?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_69684a\": \"The user's request concerns pending transfer.\", \"candidate_dbe952\": \"The user's request concerns card about to expire.\", \"candidate_29e563\": \"The user's request concerns card payment fee charged.\", \"candidate_18f27e\": \"The user's request concerns transaction charged twice.\", \"candidate_252dd7\": \"The user's request concerns pin blocked.\", \"candidate_e249df\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_8e90ea\": \"The user's request concerns passcode forgotten.\", \"candidate_c9b8cc\": \"The user's request concerns request refund.\"}}}}", "targets_json": "{\"decision\": {\"candidate_69684a\": 0.0, \"candidate_dbe952\": 0.0, \"candidate_29e563\": 1.0, \"candidate_18f27e\": 0.0, \"candidate_252dd7\": 0.0, \"candidate_e249df\": 0.0, \"candidate_8e90ea\": 0.0, \"candidate_c9b8cc\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1014d541707a699a0149df20:choice", "group_id": "fresh12:banking77:1014d541707a699a0149df20", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I payed with a card and was charged an extra fee\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d5a7a8\": \"The user's request concerns disposable card limits.\", \"candidate_18891b\": \"The user's request concerns reverted card payment?.\", \"candidate_b06ff4\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_31fce2\": \"The user's request concerns card payment fee charged.\", \"candidate_ce9a80\": \"The user's request concerns getting virtual card.\", \"candidate_5bb72e\": \"The user's request concerns atm support.\", \"candidate_4524f3\": \"The user's request concerns pin blocked.\", \"candidate_6a9458\": \"The user's request concerns cancel transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d5a7a8\": 0.0, \"candidate_18891b\": 0.0, \"candidate_b06ff4\": 0.0, \"candidate_31fce2\": 1.0, \"candidate_ce9a80\": 0.0, \"candidate_5bb72e\": 0.0, \"candidate_4524f3\": 0.0, \"candidate_6a9458\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:494b81363d51199cd0be03e4:choice", "group_id": "fresh12:banking77:494b81363d51199cd0be03e4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why do you charge fees on card payments?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_441227\": \"The user's request concerns card payment fee charged.\", \"candidate_33dc03\": \"The user's request concerns transfer timing.\", \"candidate_f9f9b1\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_d9ea89\": \"The user's request concerns cancel transfer.\", \"candidate_8a3166\": \"The user's request concerns virtual card not working.\", \"candidate_a2c5d4\": \"The user's request concerns pending transfer.\", \"candidate_b18440\": \"The user's request concerns extra charge on statement.\", \"candidate_904baa\": \"The user's request concerns top up failed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_441227\": 1.0, \"candidate_33dc03\": 0.0, \"candidate_f9f9b1\": 0.0, \"candidate_d9ea89\": 0.0, \"candidate_8a3166\": 0.0, \"candidate_a2c5d4\": 0.0, \"candidate_b18440\": 0.0, \"candidate_904baa\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:efc91701a58eedf49c20649a:choice", "group_id": "fresh12:banking77:efc91701a58eedf49c20649a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"They charged me for paying with my card.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_014725\": \"The user's request concerns disposable card limits.\", \"candidate_f93951\": \"The user's request concerns verify source of funds.\", \"candidate_136e53\": \"The user's request concerns order physical card.\", \"candidate_ae46ba\": \"The user's request concerns transfer fee charged.\", \"candidate_58ea94\": \"The user's request concerns cancel transfer.\", \"candidate_ad9a9a\": \"The user's request concerns receiving money.\", \"candidate_21bcd0\": \"The user's request concerns cash withdrawal charge.\", \"candidate_7b45f2\": \"The user's request concerns card payment fee charged.\"}}}}", "targets_json": "{\"decision\": {\"candidate_014725\": 0.0, \"candidate_f93951\": 0.0, \"candidate_136e53\": 0.0, \"candidate_ae46ba\": 0.0, \"candidate_58ea94\": 0.0, \"candidate_ad9a9a\": 0.0, \"candidate_21bcd0\": 0.0, \"candidate_7b45f2\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3a18521a89c498ab6626364c:choice", "group_id": "fresh12:banking77:3a18521a89c498ab6626364c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why was I charged an extra fee when using a card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_29d9ad\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_a40c89\": \"The user's request concerns cash withdrawal charge.\", \"candidate_8bc82b\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_d2298c\": \"The user's request concerns top up failed.\", \"candidate_2b40c1\": \"The user's request concerns compromised card.\", \"candidate_b4f557\": \"The user's request concerns card payment fee charged.\", \"candidate_251f89\": \"The user's request concerns top up by cash or cheque.\", \"candidate_5fb599\": \"The user's request concerns card delivery estimate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_29d9ad\": 0.0, \"candidate_a40c89\": 0.0, \"candidate_8bc82b\": 0.0, \"candidate_d2298c\": 0.0, \"candidate_2b40c1\": 0.0, \"candidate_b4f557\": 1.0, \"candidate_251f89\": 0.0, \"candidate_5fb599\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5dcf388df44da6bee84e46bf:choice", "group_id": "fresh12:banking77:5dcf388df44da6bee84e46bf", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is an unauthorized fee.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3fd478\": \"The user's request concerns terminate account.\", \"candidate_2c4bd8\": \"The user's request concerns automatic top up.\", \"candidate_77fe67\": \"The user's request concerns declined card payment.\", \"candidate_9de8f3\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_c5947d\": \"The user's request concerns card payment fee charged.\", \"candidate_2290e5\": \"The user's request concerns failed transfer.\", \"candidate_64a14f\": \"The user's request concerns extra charge on statement.\", \"candidate_43924d\": \"The user's request concerns topping up by card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3fd478\": 0.0, \"candidate_2c4bd8\": 0.0, \"candidate_77fe67\": 0.0, \"candidate_9de8f3\": 0.0, \"candidate_c5947d\": 1.0, \"candidate_2290e5\": 0.0, \"candidate_64a14f\": 0.0, \"candidate_43924d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ddbb4afa482835aa4cbfead1:choice", "group_id": "fresh12:banking77:ddbb4afa482835aa4cbfead1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where is the transaction I made to a friend?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dc4c30\": \"The user's request concerns extra charge on statement.\", \"candidate_8d1752\": \"The user's request concerns card payment not recognised.\", \"candidate_db91a3\": \"The user's request concerns wrong amount of cash received.\", \"candidate_41c2be\": \"The user's request concerns pending transfer.\", \"candidate_5d2f77\": \"The user's request concerns transfer not received by recipient.\", \"candidate_f3179c\": \"The user's request concerns atm support.\", \"candidate_b15435\": \"The user's request concerns card payment fee charged.\", \"candidate_49145f\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dc4c30\": 0.0, \"candidate_8d1752\": 0.0, \"candidate_db91a3\": 0.0, \"candidate_41c2be\": 0.0, \"candidate_5d2f77\": 1.0, \"candidate_f3179c\": 0.0, \"candidate_b15435\": 0.0, \"candidate_49145f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7f34b88132c42d4d04927cbf:choice", "group_id": "fresh12:banking77:7f34b88132c42d4d04927cbf", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I transferred some money and it didint arrive\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c6e7fc\": \"The user's request concerns transfer not received by recipient.\", \"candidate_ee2c4f\": \"The user's request concerns pending card payment.\", \"candidate_0d0d09\": \"The user's request concerns automatic top up.\", \"candidate_c0a0e9\": \"The user's request concerns card linking.\", \"candidate_64f8e5\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_a23a72\": \"The user's request concerns request refund.\", \"candidate_00dbdb\": \"The user's request concerns fiat currency support.\", \"candidate_683e9d\": \"The user's request concerns pending cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c6e7fc\": 1.0, \"candidate_ee2c4f\": 0.0, \"candidate_0d0d09\": 0.0, \"candidate_c0a0e9\": 0.0, \"candidate_64f8e5\": 0.0, \"candidate_a23a72\": 0.0, \"candidate_00dbdb\": 0.0, \"candidate_683e9d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b316217eb624921b134699a1:choice", "group_id": "fresh12:banking77:b316217eb624921b134699a1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long does it take for transfers to finish?  I sent funds to my pal, and she says that she has not gotten anything.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_208a16\": \"The user's request concerns terminate account.\", \"candidate_977e2d\": \"The user's request concerns declined transfer.\", \"candidate_e6abad\": \"The user's request concerns pending cash withdrawal.\", \"candidate_2c7dd7\": \"The user's request concerns transfer not received by recipient.\", \"candidate_21df42\": \"The user's request concerns fiat currency support.\", \"candidate_dd514e\": \"The user's request concerns change pin.\", \"candidate_6864f6\": \"The user's request concerns cancel transfer.\", \"candidate_2e0cd8\": \"The user's request concerns verify source of funds.\"}}}}", "targets_json": "{\"decision\": {\"candidate_208a16\": 0.0, \"candidate_977e2d\": 0.0, \"candidate_e6abad\": 0.0, \"candidate_2c7dd7\": 1.0, \"candidate_21df42\": 0.0, \"candidate_dd514e\": 0.0, \"candidate_6864f6\": 0.0, \"candidate_2e0cd8\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b41de62936292c5e1a1543a1:choice", "group_id": "fresh12:banking77:b41de62936292c5e1a1543a1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Hey, I need some help figuring out way my transfer is not complete yet. My Landlord hasn't got the money yet.She should have received it by now.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ad7948\": \"The user's request concerns age limit.\", \"candidate_e83bf0\": \"The user's request concerns beneficiary not allowed.\", \"candidate_91cc7a\": \"The user's request concerns wrong amount of cash received.\", \"candidate_9b2f6d\": \"The user's request concerns visa or mastercard.\", \"candidate_1980c6\": \"The user's request concerns transfer not received by recipient.\", \"candidate_645d38\": \"The user's request concerns lost or stolen card.\", \"candidate_0f9eec\": \"The user's request concerns cash withdrawal charge.\", \"candidate_1ef067\": \"The user's request concerns top up by cash or cheque.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ad7948\": 0.0, \"candidate_e83bf0\": 0.0, \"candidate_91cc7a\": 0.0, \"candidate_9b2f6d\": 0.0, \"candidate_1980c6\": 1.0, \"candidate_645d38\": 0.0, \"candidate_0f9eec\": 0.0, \"candidate_1ef067\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f26070c23a79e60fb5d7950f:choice", "group_id": "fresh12:banking77:f26070c23a79e60fb5d7950f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I sent some money and the receiver can't access it.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7be775\": \"The user's request concerns cancel transfer.\", \"candidate_57f76a\": \"The user's request concerns top up by cash or cheque.\", \"candidate_d5ca83\": \"The user's request concerns pending transfer.\", \"candidate_97f01a\": \"The user's request concerns age limit.\", \"candidate_6491d8\": \"The user's request concerns pin blocked.\", \"candidate_50787c\": \"The user's request concerns transfer not received by recipient.\", \"candidate_b0dfc2\": \"The user's request concerns unable to verify identity.\", \"candidate_1908c0\": \"The user's request concerns top up limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7be775\": 0.0, \"candidate_57f76a\": 0.0, \"candidate_d5ca83\": 0.0, \"candidate_97f01a\": 0.0, \"candidate_6491d8\": 0.0, \"candidate_50787c\": 1.0, \"candidate_b0dfc2\": 0.0, \"candidate_1908c0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d9c48def4cd6562ab60f4204:choice", "group_id": "fresh12:banking77:d9c48def4cd6562ab60f4204", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long will it take for my transaction to be completed?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3ef48e\": \"The user's request concerns terminate account.\", \"candidate_def104\": \"The user's request concerns card payment fee charged.\", \"candidate_a36046\": \"The user's request concerns transfer not received by recipient.\", \"candidate_9df0ae\": \"The user's request concerns pending card payment.\", \"candidate_a7da95\": \"The user's request concerns fiat currency support.\", \"candidate_36501b\": \"The user's request concerns disposable card limits.\", \"candidate_8f5762\": \"The user's request concerns verify top up.\", \"candidate_8f7d05\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3ef48e\": 0.0, \"candidate_def104\": 0.0, \"candidate_a36046\": 1.0, \"candidate_9df0ae\": 0.0, \"candidate_a7da95\": 0.0, \"candidate_36501b\": 0.0, \"candidate_8f5762\": 0.0, \"candidate_8f7d05\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:799b5184fb6b5251120577d1:choice", "group_id": "fresh12:banking77:799b5184fb6b5251120577d1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I just have one other card from the US. Is that okay?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ec1303\": \"The user's request concerns exchange rate.\", \"candidate_b9dc0a\": \"The user's request concerns receiving money.\", \"candidate_0ef78f\": \"The user's request concerns change pin.\", \"candidate_172c2f\": \"The user's request concerns Refund not showing up.\", \"candidate_e22ed3\": \"The user's request concerns reverted card payment?.\", \"candidate_23f89c\": \"The user's request concerns supported cards and currencies.\", \"candidate_cfb2dd\": \"The user's request concerns transfer fee charged.\", \"candidate_cb97a7\": \"The user's request concerns pending cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ec1303\": 0.0, \"candidate_b9dc0a\": 0.0, \"candidate_0ef78f\": 0.0, \"candidate_172c2f\": 0.0, \"candidate_e22ed3\": 0.0, \"candidate_23f89c\": 1.0, \"candidate_cfb2dd\": 0.0, \"candidate_cb97a7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9bc5b10a6d32ca077513220e:choice", "group_id": "fresh12:banking77:9bc5b10a6d32ca077513220e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What credit cards are supported?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3ea966\": \"The user's request concerns top up reverted.\", \"candidate_b1e257\": \"The user's request concerns beneficiary not allowed.\", \"candidate_bb83dd\": \"The user's request concerns declined card payment.\", \"candidate_e97b1e\": \"The user's request concerns card swallowed.\", \"candidate_2b7204\": \"The user's request concerns supported cards and currencies.\", \"candidate_197d98\": \"The user's request concerns cash withdrawal charge.\", \"candidate_a410a2\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_d2d6d0\": \"The user's request concerns automatic top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3ea966\": 0.0, \"candidate_b1e257\": 0.0, \"candidate_bb83dd\": 0.0, \"candidate_e97b1e\": 0.0, \"candidate_2b7204\": 1.0, \"candidate_197d98\": 0.0, \"candidate_a410a2\": 0.0, \"candidate_d2d6d0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4ee7c97f98abc93392e08467:choice", "group_id": "fresh12:banking77:4ee7c97f98abc93392e08467", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is my currency okay to add money with?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3e75f6\": \"The user's request concerns card arrival.\", \"candidate_e1b3e8\": \"The user's request concerns card payment fee charged.\", \"candidate_79e4c1\": \"The user's request concerns pending cash withdrawal.\", \"candidate_92da25\": \"The user's request concerns terminate account.\", \"candidate_9947af\": \"The user's request concerns supported cards and currencies.\", \"candidate_c75252\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_29529e\": \"The user's request concerns request refund.\", \"candidate_cbcd44\": \"The user's request concerns pending transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3e75f6\": 0.0, \"candidate_e1b3e8\": 0.0, \"candidate_79e4c1\": 0.0, \"candidate_92da25\": 0.0, \"candidate_9947af\": 1.0, \"candidate_c75252\": 0.0, \"candidate_29529e\": 0.0, \"candidate_cbcd44\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6471e8072e5306afee213cac:choice", "group_id": "fresh12:banking77:6471e8072e5306afee213cac", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What kind of currencies can I use to add money?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_05961f\": \"The user's request concerns top up by cash or cheque.\", \"candidate_f42c8c\": \"The user's request concerns supported cards and currencies.\", \"candidate_0f9b7e\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_195e8a\": \"The user's request concerns wrong amount of cash received.\", \"candidate_6a3fea\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_9e176f\": \"The user's request concerns Refund not showing up.\", \"candidate_72668a\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_70fa82\": \"The user's request concerns disposable card limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_05961f\": 0.0, \"candidate_f42c8c\": 1.0, \"candidate_0f9b7e\": 0.0, \"candidate_195e8a\": 0.0, \"candidate_6a3fea\": 0.0, \"candidate_9e176f\": 0.0, \"candidate_72668a\": 0.0, \"candidate_70fa82\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:50e19ce703c9e65d9f11ad29:choice", "group_id": "fresh12:banking77:50e19ce703c9e65d9f11ad29", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"To add money to my account, what currencies can I use?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3e33dc\": \"The user's request concerns disposable card limits.\", \"candidate_9b3114\": \"The user's request concerns extra charge on statement.\", \"candidate_acd613\": \"The user's request concerns unable to verify identity.\", \"candidate_93ed80\": \"The user's request concerns passcode forgotten.\", \"candidate_267b82\": \"The user's request concerns supported cards and currencies.\", \"candidate_f1ea9e\": \"The user's request concerns transaction charged twice.\", \"candidate_4513bd\": \"The user's request concerns edit personal details.\", \"candidate_3bbcab\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3e33dc\": 0.0, \"candidate_9b3114\": 0.0, \"candidate_acd613\": 0.0, \"candidate_93ed80\": 0.0, \"candidate_267b82\": 1.0, \"candidate_f1ea9e\": 0.0, \"candidate_4513bd\": 0.0, \"candidate_3bbcab\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4e56a54475b304a784622e28:choice", "group_id": "fresh12:banking77:4e56a54475b304a784622e28", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I use American Express to add money into my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_78716d\": \"The user's request concerns request refund.\", \"candidate_e42a17\": \"The user's request concerns country support.\", \"candidate_52877d\": \"The user's request concerns cash withdrawal charge.\", \"candidate_4f3e2f\": \"The user's request concerns supported cards and currencies.\", \"candidate_2496d4\": \"The user's request concerns card swallowed.\", \"candidate_42257f\": \"The user's request concerns pending cash withdrawal.\", \"candidate_4614d0\": \"The user's request concerns automatic top up.\", \"candidate_d3e471\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_78716d\": 0.0, \"candidate_e42a17\": 0.0, \"candidate_52877d\": 0.0, \"candidate_4f3e2f\": 1.0, \"candidate_2496d4\": 0.0, \"candidate_42257f\": 0.0, \"candidate_4614d0\": 0.0, \"candidate_d3e471\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:20cc5d9fd6856d0086c6ce6f:choice", "group_id": "fresh12:banking77:20cc5d9fd6856d0086c6ce6f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What are the steps to getting a virtual card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e8e861\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_79ad09\": \"The user's request concerns visa or mastercard.\", \"candidate_0a8546\": \"The user's request concerns pending transfer.\", \"candidate_914e67\": \"The user's request concerns disposable card limits.\", \"candidate_222cf1\": \"The user's request concerns exchange charge.\", \"candidate_a906ff\": \"The user's request concerns pin blocked.\", \"candidate_3c8ae5\": \"The user's request concerns getting virtual card.\", \"candidate_bd575d\": \"The user's request concerns pending cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e8e861\": 0.0, \"candidate_79ad09\": 0.0, \"candidate_0a8546\": 0.0, \"candidate_914e67\": 0.0, \"candidate_222cf1\": 0.0, \"candidate_a906ff\": 0.0, \"candidate_3c8ae5\": 1.0, \"candidate_bd575d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:16a6010b4b30b1002950a100:choice", "group_id": "fresh12:banking77:16a6010b4b30b1002950a100", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"help me obtain a virtual card\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_19f155\": \"The user's request concerns pending card payment.\", \"candidate_ddc8c3\": \"The user's request concerns transfer not received by recipient.\", \"candidate_3f13f5\": \"The user's request concerns virtual card not working.\", \"candidate_eb121a\": \"The user's request concerns card about to expire.\", \"candidate_b948fd\": \"The user's request concerns card arrival.\", \"candidate_1c5e34\": \"The user's request concerns getting virtual card.\", \"candidate_ac6809\": \"The user's request concerns getting spare card.\", \"candidate_4efa8d\": \"The user's request concerns card delivery estimate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_19f155\": 0.0, \"candidate_ddc8c3\": 0.0, \"candidate_3f13f5\": 0.0, \"candidate_eb121a\": 0.0, \"candidate_b948fd\": 0.0, \"candidate_1c5e34\": 1.0, \"candidate_ac6809\": 0.0, \"candidate_4efa8d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2c13bea2c9f65de329d49c26:choice", "group_id": "fresh12:banking77:2c13bea2c9f65de329d49c26", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I order a virtual card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e15dfa\": \"The user's request concerns top up by card charge.\", \"candidate_d2c822\": \"The user's request concerns exchange via app.\", \"candidate_7d9e23\": \"The user's request concerns order physical card.\", \"candidate_7f9f19\": \"The user's request concerns topping up by card.\", \"candidate_db521a\": \"The user's request concerns getting virtual card.\", \"candidate_6f9a2c\": \"The user's request concerns beneficiary not allowed.\", \"candidate_3188c0\": \"The user's request concerns apple pay or google pay.\", \"candidate_297021\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e15dfa\": 0.0, \"candidate_d2c822\": 0.0, \"candidate_7d9e23\": 0.0, \"candidate_7f9f19\": 0.0, \"candidate_db521a\": 1.0, \"candidate_6f9a2c\": 0.0, \"candidate_3188c0\": 0.0, \"candidate_297021\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:59220bec8dfcb171b48327b0:choice", "group_id": "fresh12:banking77:59220bec8dfcb171b48327b0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I have one of the virtual cards?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cb97d4\": \"The user's request concerns unable to verify identity.\", \"candidate_0366a2\": \"The user's request concerns automatic top up.\", \"candidate_ce9c97\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_e51b1a\": \"The user's request concerns getting virtual card.\", \"candidate_3835e0\": \"The user's request concerns receiving money.\", \"candidate_71bd40\": \"The user's request concerns verify top up.\", \"candidate_6da52e\": \"The user's request concerns activate my card.\", \"candidate_fa7306\": \"The user's request concerns edit personal details.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cb97d4\": 0.0, \"candidate_0366a2\": 0.0, \"candidate_ce9c97\": 0.0, \"candidate_e51b1a\": 1.0, \"candidate_3835e0\": 0.0, \"candidate_71bd40\": 0.0, \"candidate_6da52e\": 0.0, \"candidate_fa7306\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:905144531b3a7254d75db765:choice", "group_id": "fresh12:banking77:905144531b3a7254d75db765", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there an option to have a virtual card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a0221e\": \"The user's request concerns getting virtual card.\", \"candidate_1273ed\": \"The user's request concerns top up limits.\", \"candidate_106cf3\": \"The user's request concerns verify source of funds.\", \"candidate_828495\": \"The user's request concerns pending card payment.\", \"candidate_c66232\": \"The user's request concerns top up by card charge.\", \"candidate_7480d5\": \"The user's request concerns exchange via app.\", \"candidate_237a7c\": \"The user's request concerns edit personal details.\", \"candidate_bb53c3\": \"The user's request concerns passcode forgotten.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a0221e\": 1.0, \"candidate_1273ed\": 0.0, \"candidate_106cf3\": 0.0, \"candidate_828495\": 0.0, \"candidate_c66232\": 0.0, \"candidate_7480d5\": 0.0, \"candidate_237a7c\": 0.0, \"candidate_bb53c3\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1690d797d9b3251bbb9741d4:choice", "group_id": "fresh12:banking77:1690d797d9b3251bbb9741d4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where can I obtain my virtual card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f334f1\": \"The user's request concerns topping up by card.\", \"candidate_16c160\": \"The user's request concerns terminate account.\", \"candidate_630c4d\": \"The user's request concerns card swallowed.\", \"candidate_a0a6bc\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_4e8d9b\": \"The user's request concerns getting virtual card.\", \"candidate_6d6ce7\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_b92845\": \"The user's request concerns card not working.\", \"candidate_553f64\": \"The user's request concerns getting spare card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f334f1\": 0.0, \"candidate_16c160\": 0.0, \"candidate_630c4d\": 0.0, \"candidate_a0a6bc\": 0.0, \"candidate_4e8d9b\": 1.0, \"candidate_6d6ce7\": 0.0, \"candidate_b92845\": 0.0, \"candidate_553f64\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a5d11eb385d3490beaaaf2a6:choice", "group_id": "fresh12:banking77:a5d11eb385d3490beaaaf2a6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What are the rules to where I can use my card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3c1b3f\": \"The user's request concerns exchange charge.\", \"candidate_4cdf75\": \"The user's request concerns extra charge on statement.\", \"candidate_ec6c17\": \"The user's request concerns card acceptance.\", \"candidate_d8c20e\": \"The user's request concerns transfer into account.\", \"candidate_96a610\": \"The user's request concerns unable to verify identity.\", \"candidate_4222db\": \"The user's request concerns top up failed.\", \"candidate_d86975\": \"The user's request concerns card not working.\", \"candidate_2301f1\": \"The user's request concerns verify top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3c1b3f\": 0.0, \"candidate_4cdf75\": 0.0, \"candidate_ec6c17\": 1.0, \"candidate_d8c20e\": 0.0, \"candidate_96a610\": 0.0, \"candidate_4222db\": 0.0, \"candidate_d86975\": 0.0, \"candidate_2301f1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1226110c45ffc0eeb8929cc1:choice", "group_id": "fresh12:banking77:1226110c45ffc0eeb8929cc1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"places i can use the card\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3457fa\": \"The user's request concerns pending cash withdrawal.\", \"candidate_37232c\": \"The user's request concerns Refund not showing up.\", \"candidate_92f264\": \"The user's request concerns get disposable virtual card.\", \"candidate_d8276a\": \"The user's request concerns card arrival.\", \"candidate_080437\": \"The user's request concerns extra charge on statement.\", \"candidate_50f7b2\": \"The user's request concerns cash withdrawal charge.\", \"candidate_2ffe1d\": \"The user's request concerns card acceptance.\", \"candidate_b565b9\": \"The user's request concerns card swallowed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3457fa\": 0.0, \"candidate_37232c\": 0.0, \"candidate_92f264\": 0.0, \"candidate_d8276a\": 0.0, \"candidate_080437\": 0.0, \"candidate_50f7b2\": 0.0, \"candidate_2ffe1d\": 1.0, \"candidate_b565b9\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:89280aef2f9d1c35535bf96e:choice", "group_id": "fresh12:banking77:89280aef2f9d1c35535bf96e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where can I pay with my debit or credit card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b06268\": \"The user's request concerns wrong amount of cash received.\", \"candidate_09af31\": \"The user's request concerns card delivery estimate.\", \"candidate_964f82\": \"The user's request concerns top up by card charge.\", \"candidate_55cddf\": \"The user's request concerns compromised card.\", \"candidate_2e8d1a\": \"The user's request concerns card acceptance.\", \"candidate_0ff580\": \"The user's request concerns top up reverted.\", \"candidate_112967\": \"The user's request concerns card not working.\", \"candidate_e50d4e\": \"The user's request concerns exchange via app.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b06268\": 0.0, \"candidate_09af31\": 0.0, \"candidate_964f82\": 0.0, \"candidate_55cddf\": 0.0, \"candidate_2e8d1a\": 1.0, \"candidate_0ff580\": 0.0, \"candidate_112967\": 0.0, \"candidate_e50d4e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7d90ba4ddbcb5b39ad8dd8c9:choice", "group_id": "fresh12:banking77:7d90ba4ddbcb5b39ad8dd8c9", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What stores can I sue my card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_04c197\": \"The user's request concerns disposable card limits.\", \"candidate_c1a6fe\": \"The user's request concerns card acceptance.\", \"candidate_877fb9\": \"The user's request concerns declined cash withdrawal.\", \"candidate_afc3df\": \"The user's request concerns top up limits.\", \"candidate_0a011d\": \"The user's request concerns passcode forgotten.\", \"candidate_9df465\": \"The user's request concerns request refund.\", \"candidate_f534f5\": \"The user's request concerns verify top up.\", \"candidate_c8b913\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_04c197\": 0.0, \"candidate_c1a6fe\": 1.0, \"candidate_877fb9\": 0.0, \"candidate_afc3df\": 0.0, \"candidate_0a011d\": 0.0, \"candidate_9df465\": 0.0, \"candidate_f534f5\": 0.0, \"candidate_c8b913\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d79cf0dd7c37ffc2b1c02960:choice", "group_id": "fresh12:banking77:d79cf0dd7c37ffc2b1c02960", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where is my card accepted?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f932f4\": \"The user's request concerns activate my card.\", \"candidate_d2521f\": \"The user's request concerns card acceptance.\", \"candidate_91f50a\": \"The user's request concerns Refund not showing up.\", \"candidate_140a7e\": \"The user's request concerns failed transfer.\", \"candidate_afabf1\": \"The user's request concerns extra charge on statement.\", \"candidate_aef5e2\": \"The user's request concerns pending top up.\", \"candidate_1ce2eb\": \"The user's request concerns cancel transfer.\", \"candidate_f94809\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f932f4\": 0.0, \"candidate_d2521f\": 1.0, \"candidate_91f50a\": 0.0, \"candidate_140a7e\": 0.0, \"candidate_afabf1\": 0.0, \"candidate_aef5e2\": 0.0, \"candidate_1ce2eb\": 0.0, \"candidate_f94809\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:36c5c1031f97cc3ce6f340e2:choice", "group_id": "fresh12:banking77:36c5c1031f97cc3ce6f340e2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I use my card everywhere?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3a39ff\": \"The user's request concerns card acceptance.\", \"candidate_f4c0e7\": \"The user's request concerns exchange charge.\", \"candidate_183357\": \"The user's request concerns getting spare card.\", \"candidate_8e218a\": \"The user's request concerns pending card payment.\", \"candidate_0762b5\": \"The user's request concerns apple pay or google pay.\", \"candidate_e5f4b3\": \"The user's request concerns pending transfer.\", \"candidate_665b30\": \"The user's request concerns automatic top up.\", \"candidate_c36022\": \"The user's request concerns lost or stolen card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3a39ff\": 1.0, \"candidate_f4c0e7\": 0.0, \"candidate_183357\": 0.0, \"candidate_8e218a\": 0.0, \"candidate_0762b5\": 0.0, \"candidate_e5f4b3\": 0.0, \"candidate_665b30\": 0.0, \"candidate_c36022\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c0001ec6de1577a47d14fcb2:choice", "group_id": "fresh12:banking77:c0001ec6de1577a47d14fcb2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"On my last transaction it seem that my top-up was not successful.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5124ba\": \"The user's request concerns topping up by card.\", \"candidate_87ce0f\": \"The user's request concerns getting virtual card.\", \"candidate_5e2d6c\": \"The user's request concerns top up reverted.\", \"candidate_35d243\": \"The user's request concerns card payment not recognised.\", \"candidate_3ff9f8\": \"The user's request concerns activate my card.\", \"candidate_a2e6e2\": \"The user's request concerns visa or mastercard.\", \"candidate_dd9c0a\": \"The user's request concerns card payment fee charged.\", \"candidate_ac696f\": \"The user's request concerns request refund.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5124ba\": 0.0, \"candidate_87ce0f\": 0.0, \"candidate_5e2d6c\": 1.0, \"candidate_35d243\": 0.0, \"candidate_3ff9f8\": 0.0, \"candidate_a2e6e2\": 0.0, \"candidate_dd9c0a\": 0.0, \"candidate_ac696f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6fdab92ff969c55a51b9fac6:choice", "group_id": "fresh12:banking77:6fdab92ff969c55a51b9fac6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I put money into my account for the minimum balance but the application didn't accept.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7665e1\": \"The user's request concerns top up reverted.\", \"candidate_ccaba4\": \"The user's request concerns supported cards and currencies.\", \"candidate_367338\": \"The user's request concerns virtual card not working.\", \"candidate_5032e4\": \"The user's request concerns apple pay or google pay.\", \"candidate_ace769\": \"The user's request concerns verify my identity.\", \"candidate_586a37\": \"The user's request concerns get disposable virtual card.\", \"candidate_744aae\": \"The user's request concerns contactless not working.\", \"candidate_6f996f\": \"The user's request concerns exchange via app.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7665e1\": 1.0, \"candidate_ccaba4\": 0.0, \"candidate_367338\": 0.0, \"candidate_5032e4\": 0.0, \"candidate_ace769\": 0.0, \"candidate_586a37\": 0.0, \"candidate_744aae\": 0.0, \"candidate_6f996f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7b3d8c2d3ea91ab02122cbc8:choice", "group_id": "fresh12:banking77:7b3d8c2d3ea91ab02122cbc8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I'd like to know why my top-up was canceled please.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c356dc\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_85b948\": \"The user's request concerns pending transfer.\", \"candidate_9f7679\": \"The user's request concerns top up reverted.\", \"candidate_749a03\": \"The user's request concerns top up failed.\", \"candidate_ece31c\": \"The user's request concerns request refund.\", \"candidate_491d9d\": \"The user's request concerns top up limits.\", \"candidate_b93296\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_fc203a\": \"The user's request concerns balance not updated after bank transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c356dc\": 0.0, \"candidate_85b948\": 0.0, \"candidate_9f7679\": 1.0, \"candidate_749a03\": 0.0, \"candidate_ece31c\": 0.0, \"candidate_491d9d\": 0.0, \"candidate_b93296\": 0.0, \"candidate_fc203a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1cc523e57b92b0b295c3383a:choice", "group_id": "fresh12:banking77:1cc523e57b92b0b295c3383a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why would my top up be cancelled?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_53cf4e\": \"The user's request concerns topping up by card.\", \"candidate_43565c\": \"The user's request concerns compromised card.\", \"candidate_1927be\": \"The user's request concerns top up reverted.\", \"candidate_380042\": \"The user's request concerns verify source of funds.\", \"candidate_b6332b\": \"The user's request concerns card not working.\", \"candidate_4e789d\": \"The user's request concerns card linking.\", \"candidate_50bd39\": \"The user's request concerns passcode forgotten.\", \"candidate_967771\": \"The user's request concerns wrong exchange rate for cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_53cf4e\": 0.0, \"candidate_43565c\": 0.0, \"candidate_1927be\": 1.0, \"candidate_380042\": 0.0, \"candidate_b6332b\": 0.0, \"candidate_4e789d\": 0.0, \"candidate_50bd39\": 0.0, \"candidate_967771\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ca0899cd57a8352d8f7475c8:choice", "group_id": "fresh12:banking77:ca0899cd57a8352d8f7475c8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I fix a reverted top up?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d3102a\": \"The user's request concerns top up reverted.\", \"candidate_74a91e\": \"The user's request concerns declined cash withdrawal.\", \"candidate_b832fa\": \"The user's request concerns atm support.\", \"candidate_d4f4ee\": \"The user's request concerns compromised card.\", \"candidate_4e3c1d\": \"The user's request concerns receiving money.\", \"candidate_177979\": \"The user's request concerns getting virtual card.\", \"candidate_1d5614\": \"The user's request concerns unable to verify identity.\", \"candidate_30a516\": \"The user's request concerns fiat currency support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d3102a\": 1.0, \"candidate_74a91e\": 0.0, \"candidate_b832fa\": 0.0, \"candidate_d4f4ee\": 0.0, \"candidate_4e3c1d\": 0.0, \"candidate_177979\": 0.0, \"candidate_1d5614\": 0.0, \"candidate_30a516\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f49e98a6c3243df0396f8949:choice", "group_id": "fresh12:banking77:f49e98a6c3243df0396f8949", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I believe my money did not go through with my top up, was there a problem on your end?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9cf091\": \"The user's request concerns transfer into account.\", \"candidate_41fe7d\": \"The user's request concerns card not working.\", \"candidate_f56e2b\": \"The user's request concerns pending card payment.\", \"candidate_ef63ca\": \"The user's request concerns top up reverted.\", \"candidate_ca0c7f\": \"The user's request concerns extra charge on statement.\", \"candidate_b0a0c9\": \"The user's request concerns visa or mastercard.\", \"candidate_64ec2b\": \"The user's request concerns transfer not received by recipient.\", \"candidate_e043ca\": \"The user's request concerns top up by bank transfer charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9cf091\": 0.0, \"candidate_41fe7d\": 0.0, \"candidate_f56e2b\": 0.0, \"candidate_ef63ca\": 1.0, \"candidate_ca0c7f\": 0.0, \"candidate_b0a0c9\": 0.0, \"candidate_64ec2b\": 0.0, \"candidate_e043ca\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5d9c4eb896064c0828a36e7a:choice", "group_id": "fresh12:banking77:5d9c4eb896064c0828a36e7a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I made a deposit of cash and check that has not been added to my balance.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dad5c9\": \"The user's request concerns reverted card payment?.\", \"candidate_fc0a99\": \"The user's request concerns verify top up.\", \"candidate_6e4769\": \"The user's request concerns card arrival.\", \"candidate_9ea305\": \"The user's request concerns verify my identity.\", \"candidate_f896cb\": \"The user's request concerns card payment not recognised.\", \"candidate_7477b0\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_082a4b\": \"The user's request concerns pending cash withdrawal.\", \"candidate_d0e446\": \"The user's request concerns card acceptance.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dad5c9\": 0.0, \"candidate_fc0a99\": 0.0, \"candidate_6e4769\": 0.0, \"candidate_9ea305\": 0.0, \"candidate_f896cb\": 0.0, \"candidate_7477b0\": 1.0, \"candidate_082a4b\": 0.0, \"candidate_d0e446\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f209693185de588289cfeb67:choice", "group_id": "fresh12:banking77:f209693185de588289cfeb67", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How come my cash deposit is not showing up. This can't be right. There must be some mistake. Where did my cash go. You better not have lost it as I need this money asap.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c89b3c\": \"The user's request concerns verify my identity.\", \"candidate_c2605a\": \"The user's request concerns cancel transfer.\", \"candidate_3166db\": \"The user's request concerns transfer not received by recipient.\", \"candidate_94b747\": \"The user's request concerns supported cards and currencies.\", \"candidate_0abb24\": \"The user's request concerns passcode forgotten.\", \"candidate_f63e9b\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_64bbb8\": \"The user's request concerns age limit.\", \"candidate_ff9dec\": \"The user's request concerns pending cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c89b3c\": 0.0, \"candidate_c2605a\": 0.0, \"candidate_3166db\": 0.0, \"candidate_94b747\": 0.0, \"candidate_0abb24\": 0.0, \"candidate_f63e9b\": 1.0, \"candidate_64bbb8\": 0.0, \"candidate_ff9dec\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5b6cc0c3f0c2ada68fae7ef6:choice", "group_id": "fresh12:banking77:5b6cc0c3f0c2ada68fae7ef6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Cash deposit to my account but it has not appeared.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_73d066\": \"The user's request concerns get disposable virtual card.\", \"candidate_151271\": \"The user's request concerns receiving money.\", \"candidate_2b5fac\": \"The user's request concerns visa or mastercard.\", \"candidate_705931\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_d6f978\": \"The user's request concerns transfer into account.\", \"candidate_86c13b\": \"The user's request concerns cash withdrawal charge.\", \"candidate_b6f5cd\": \"The user's request concerns supported cards and currencies.\", \"candidate_a8828f\": \"The user's request concerns balance not updated after cheque or cash deposit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_73d066\": 0.0, \"candidate_151271\": 0.0, \"candidate_2b5fac\": 0.0, \"candidate_705931\": 0.0, \"candidate_d6f978\": 0.0, \"candidate_86c13b\": 0.0, \"candidate_b6f5cd\": 0.0, \"candidate_a8828f\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f572a3b0e1bf406baeccc2eb:choice", "group_id": "fresh12:banking77:f572a3b0e1bf406baeccc2eb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why does my account not accept cash deposits?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3b1b65\": \"The user's request concerns beneficiary not allowed.\", \"candidate_8e025c\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_a9c5f8\": \"The user's request concerns verify my identity.\", \"candidate_6721d3\": \"The user's request concerns exchange charge.\", \"candidate_75b0e3\": \"The user's request concerns unable to verify identity.\", \"candidate_a99a50\": \"The user's request concerns declined cash withdrawal.\", \"candidate_6370c2\": \"The user's request concerns transfer not received by recipient.\", \"candidate_07824b\": \"The user's request concerns balance not updated after cheque or cash deposit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3b1b65\": 0.0, \"candidate_8e025c\": 0.0, \"candidate_a9c5f8\": 0.0, \"candidate_6721d3\": 0.0, \"candidate_75b0e3\": 0.0, \"candidate_a99a50\": 0.0, \"candidate_6370c2\": 0.0, \"candidate_07824b\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:44ae84827fbd7f0f36bb06a3:choice", "group_id": "fresh12:banking77:44ae84827fbd7f0f36bb06a3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Balance hasn't been updated following a cheque or cash deposit\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4e98ad\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_c12c91\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_1be0fb\": \"The user's request concerns declined cash withdrawal.\", \"candidate_5fc28a\": \"The user's request concerns pin blocked.\", \"candidate_3c3f75\": \"The user's request concerns transfer fee charged.\", \"candidate_37fa4a\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_71e5ae\": \"The user's request concerns beneficiary not allowed.\", \"candidate_d9d90c\": \"The user's request concerns card swallowed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4e98ad\": 0.0, \"candidate_c12c91\": 1.0, \"candidate_1be0fb\": 0.0, \"candidate_5fc28a\": 0.0, \"candidate_3c3f75\": 0.0, \"candidate_37fa4a\": 0.0, \"candidate_71e5ae\": 0.0, \"candidate_d9d90c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:386668479b30b58412b59ddb:choice", "group_id": "fresh12:banking77:386668479b30b58412b59ddb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I deposited cash this morning and I am still waiting\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_462bea\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_eb29a7\": \"The user's request concerns declined transfer.\", \"candidate_2939c3\": \"The user's request concerns exchange rate.\", \"candidate_6bd276\": \"The user's request concerns get physical card.\", \"candidate_861e54\": \"The user's request concerns apple pay or google pay.\", \"candidate_42c354\": \"The user's request concerns reverted card payment?.\", \"candidate_07be75\": \"The user's request concerns getting spare card.\", \"candidate_a54552\": \"The user's request concerns receiving money.\"}}}}", "targets_json": "{\"decision\": {\"candidate_462bea\": 1.0, \"candidate_eb29a7\": 0.0, \"candidate_2939c3\": 0.0, \"candidate_6bd276\": 0.0, \"candidate_861e54\": 0.0, \"candidate_42c354\": 0.0, \"candidate_07be75\": 0.0, \"candidate_a54552\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:02e43d78b606e67f78db17ed:choice", "group_id": "fresh12:banking77:02e43d78b606e67f78db17ed", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"there is a transaction on my account that i didnt make\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c56005\": \"The user's request concerns lost or stolen card.\", \"candidate_03d6d1\": \"The user's request concerns card payment not recognised.\", \"candidate_a75646\": \"The user's request concerns transfer into account.\", \"candidate_589d0a\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_2939f5\": \"The user's request concerns card arrival.\", \"candidate_253858\": \"The user's request concerns activate my card.\", \"candidate_49b87c\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_e3b745\": \"The user's request concerns receiving money.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c56005\": 0.0, \"candidate_03d6d1\": 1.0, \"candidate_a75646\": 0.0, \"candidate_589d0a\": 0.0, \"candidate_2939f5\": 0.0, \"candidate_253858\": 0.0, \"candidate_49b87c\": 0.0, \"candidate_e3b745\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e9b5a4b9ab43ae73f9508503:choice", "group_id": "fresh12:banking77:e9b5a4b9ab43ae73f9508503", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I did not make this card payment.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5f475a\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_d164c4\": \"The user's request concerns compromised card.\", \"candidate_991f17\": \"The user's request concerns card delivery estimate.\", \"candidate_afefb4\": \"The user's request concerns pin blocked.\", \"candidate_b99302\": \"The user's request concerns card not working.\", \"candidate_89a0f4\": \"The user's request concerns reverted card payment?.\", \"candidate_7a493b\": \"The user's request concerns card payment not recognised.\", \"candidate_a410e9\": \"The user's request concerns card acceptance.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5f475a\": 0.0, \"candidate_d164c4\": 0.0, \"candidate_991f17\": 0.0, \"candidate_afefb4\": 0.0, \"candidate_b99302\": 0.0, \"candidate_89a0f4\": 0.0, \"candidate_7a493b\": 1.0, \"candidate_a410e9\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5452228ce1db61f073c12995:choice", "group_id": "fresh12:banking77:5452228ce1db61f073c12995", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is a payment in the App that is not mine.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d65bd7\": \"The user's request concerns fiat currency support.\", \"candidate_2cf23d\": \"The user's request concerns verify my identity.\", \"candidate_126fdd\": \"The user's request concerns get disposable virtual card.\", \"candidate_a36f42\": \"The user's request concerns terminate account.\", \"candidate_efc4f5\": \"The user's request concerns request refund.\", \"candidate_24243b\": \"The user's request concerns card payment not recognised.\", \"candidate_7cb468\": \"The user's request concerns topping up by card.\", \"candidate_f8086f\": \"The user's request concerns supported cards and currencies.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d65bd7\": 0.0, \"candidate_2cf23d\": 0.0, \"candidate_126fdd\": 0.0, \"candidate_a36f42\": 0.0, \"candidate_efc4f5\": 0.0, \"candidate_24243b\": 1.0, \"candidate_7cb468\": 0.0, \"candidate_f8086f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7b8eed6d1754f580747b407b:choice", "group_id": "fresh12:banking77:7b8eed6d1754f580747b407b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I have an unauthorized transaction on my statement\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bcf0ad\": \"The user's request concerns activate my card.\", \"candidate_a2c97f\": \"The user's request concerns pending cash withdrawal.\", \"candidate_d6740c\": \"The user's request concerns card about to expire.\", \"candidate_8de96b\": \"The user's request concerns exchange via app.\", \"candidate_97fdd1\": \"The user's request concerns transaction charged twice.\", \"candidate_1780b4\": \"The user's request concerns pending top up.\", \"candidate_007da5\": \"The user's request concerns verify my identity.\", \"candidate_5f3f31\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bcf0ad\": 0.0, \"candidate_a2c97f\": 0.0, \"candidate_d6740c\": 0.0, \"candidate_8de96b\": 0.0, \"candidate_97fdd1\": 0.0, \"candidate_1780b4\": 0.0, \"candidate_007da5\": 0.0, \"candidate_5f3f31\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c7a994caae910c00dcf92229:choice", "group_id": "fresh12:banking77:c7a994caae910c00dcf92229", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I don't recognise a card payment\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c9851c\": \"The user's request concerns verify source of funds.\", \"candidate_6b6616\": \"The user's request concerns card arrival.\", \"candidate_7ee546\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_8feb37\": \"The user's request concerns lost or stolen phone.\", \"candidate_560748\": \"The user's request concerns terminate account.\", \"candidate_c76777\": \"The user's request concerns card payment not recognised.\", \"candidate_1120a2\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_d264b4\": \"The user's request concerns top up failed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c9851c\": 0.0, \"candidate_6b6616\": 0.0, \"candidate_7ee546\": 0.0, \"candidate_8feb37\": 0.0, \"candidate_560748\": 0.0, \"candidate_c76777\": 1.0, \"candidate_1120a2\": 0.0, \"candidate_d264b4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ce9b0d59fdb803ce08a83113:choice", "group_id": "fresh12:banking77:ce9b0d59fdb803ce08a83113", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I have a payment listed in error.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_15833a\": \"The user's request concerns get disposable virtual card.\", \"candidate_9333a7\": \"The user's request concerns change pin.\", \"candidate_dc6b63\": \"The user's request concerns pin blocked.\", \"candidate_f192f4\": \"The user's request concerns edit personal details.\", \"candidate_da6d8a\": \"The user's request concerns exchange rate.\", \"candidate_d3798e\": \"The user's request concerns transfer not received by recipient.\", \"candidate_3960d1\": \"The user's request concerns transfer into account.\", \"candidate_fe8bec\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_15833a\": 0.0, \"candidate_9333a7\": 0.0, \"candidate_dc6b63\": 0.0, \"candidate_f192f4\": 0.0, \"candidate_da6d8a\": 0.0, \"candidate_d3798e\": 0.0, \"candidate_3960d1\": 0.0, \"candidate_fe8bec\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b9cd2157490b82b41923621c:choice", "group_id": "fresh12:banking77:b9cd2157490b82b41923621c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is it possible to edit my personal details on the app?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9ab30a\": \"The user's request concerns extra charge on statement.\", \"candidate_e7b29a\": \"The user's request concerns getting spare card.\", \"candidate_fdf3e8\": \"The user's request concerns visa or mastercard.\", \"candidate_8d405c\": \"The user's request concerns receiving money.\", \"candidate_146555\": \"The user's request concerns edit personal details.\", \"candidate_bf17a0\": \"The user's request concerns top up by cash or cheque.\", \"candidate_b0d9f3\": \"The user's request concerns transfer fee charged.\", \"candidate_39614d\": \"The user's request concerns automatic top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9ab30a\": 0.0, \"candidate_e7b29a\": 0.0, \"candidate_fdf3e8\": 0.0, \"candidate_8d405c\": 0.0, \"candidate_146555\": 1.0, \"candidate_bf17a0\": 0.0, \"candidate_b0d9f3\": 0.0, \"candidate_39614d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a96e4cd83b061e7e32179957:choice", "group_id": "fresh12:banking77:a96e4cd83b061e7e32179957", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need to update my info.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bda4d9\": \"The user's request concerns request refund.\", \"candidate_d1aafc\": \"The user's request concerns declined card payment.\", \"candidate_998afa\": \"The user's request concerns pending card payment.\", \"candidate_c50654\": \"The user's request concerns edit personal details.\", \"candidate_a363ca\": \"The user's request concerns pending top up.\", \"candidate_28f3a7\": \"The user's request concerns transfer into account.\", \"candidate_b5a0dc\": \"The user's request concerns Refund not showing up.\", \"candidate_929be1\": \"The user's request concerns transfer fee charged.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bda4d9\": 0.0, \"candidate_d1aafc\": 0.0, \"candidate_998afa\": 0.0, \"candidate_c50654\": 1.0, \"candidate_a363ca\": 0.0, \"candidate_28f3a7\": 0.0, \"candidate_b5a0dc\": 0.0, \"candidate_929be1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:fda009d37b67aa239200beaa:choice", "group_id": "fresh12:banking77:fda009d37b67aa239200beaa", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where can I go to update my personal profile?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fb3f08\": \"The user's request concerns transfer not received by recipient.\", \"candidate_bb9958\": \"The user's request concerns get disposable virtual card.\", \"candidate_bf599b\": \"The user's request concerns cash withdrawal charge.\", \"candidate_cdca5e\": \"The user's request concerns edit personal details.\", \"candidate_341527\": \"The user's request concerns card payment fee charged.\", \"candidate_f8edef\": \"The user's request concerns visa or mastercard.\", \"candidate_6812ba\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_f4e7a4\": \"The user's request concerns unable to verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fb3f08\": 0.0, \"candidate_bb9958\": 0.0, \"candidate_bf599b\": 0.0, \"candidate_cdca5e\": 1.0, \"candidate_341527\": 0.0, \"candidate_f8edef\": 0.0, \"candidate_6812ba\": 0.0, \"candidate_f4e7a4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:711d406c719bbcfa46617519:choice", "group_id": "fresh12:banking77:711d406c719bbcfa46617519", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need to change my account details\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2270fb\": \"The user's request concerns virtual card not working.\", \"candidate_b2e897\": \"The user's request concerns card payment fee charged.\", \"candidate_948f22\": \"The user's request concerns exchange charge.\", \"candidate_3ce7a7\": \"The user's request concerns change pin.\", \"candidate_87f907\": \"The user's request concerns activate my card.\", \"candidate_535f1c\": \"The user's request concerns card linking.\", \"candidate_f8f025\": \"The user's request concerns edit personal details.\", \"candidate_ff0a8c\": \"The user's request concerns top up by bank transfer charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2270fb\": 0.0, \"candidate_b2e897\": 0.0, \"candidate_948f22\": 0.0, \"candidate_3ce7a7\": 0.0, \"candidate_87f907\": 0.0, \"candidate_535f1c\": 0.0, \"candidate_f8f025\": 1.0, \"candidate_ff0a8c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e5b103a5b6cf1982a5a53579:choice", "group_id": "fresh12:banking77:e5b103a5b6cf1982a5a53579", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need to update my current address\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1afa20\": \"The user's request concerns card linking.\", \"candidate_e99d48\": \"The user's request concerns card swallowed.\", \"candidate_dfa127\": \"The user's request concerns edit personal details.\", \"candidate_04b8e8\": \"The user's request concerns topping up by card.\", \"candidate_99f99a\": \"The user's request concerns activate my card.\", \"candidate_b3c0a2\": \"The user's request concerns cash withdrawal charge.\", \"candidate_e39983\": \"The user's request concerns wrong amount of cash received.\", \"candidate_e1f431\": \"The user's request concerns request refund.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1afa20\": 0.0, \"candidate_e99d48\": 0.0, \"candidate_dfa127\": 1.0, \"candidate_04b8e8\": 0.0, \"candidate_99f99a\": 0.0, \"candidate_b3c0a2\": 0.0, \"candidate_e39983\": 0.0, \"candidate_e1f431\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0781add57f2f593f3696d216:choice", "group_id": "fresh12:banking77:0781add57f2f593f3696d216", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I moved.  I need to update my details.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8f106e\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_589fe3\": \"The user's request concerns card payment not recognised.\", \"candidate_224465\": \"The user's request concerns transaction charged twice.\", \"candidate_be8d27\": \"The user's request concerns atm support.\", \"candidate_91146d\": \"The user's request concerns card linking.\", \"candidate_513839\": \"The user's request concerns get physical card.\", \"candidate_1dbfbc\": \"The user's request concerns declined cash withdrawal.\", \"candidate_9cca35\": \"The user's request concerns edit personal details.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8f106e\": 0.0, \"candidate_589fe3\": 0.0, \"candidate_224465\": 0.0, \"candidate_be8d27\": 0.0, \"candidate_91146d\": 0.0, \"candidate_513839\": 0.0, \"candidate_1dbfbc\": 0.0, \"candidate_9cca35\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f79554274963622b020f8e55:choice", "group_id": "fresh12:banking77:f79554274963622b020f8e55", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What is the need to verify my identity?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4f6fce\": \"The user's request concerns card linking.\", \"candidate_931bf7\": \"The user's request concerns passcode forgotten.\", \"candidate_af633c\": \"The user's request concerns declined card payment.\", \"candidate_7ecd3c\": \"The user's request concerns why verify identity.\", \"candidate_29dbc8\": \"The user's request concerns pin blocked.\", \"candidate_11afa5\": \"The user's request concerns top up limits.\", \"candidate_bce0e3\": \"The user's request concerns receiving money.\", \"candidate_c2de1e\": \"The user's request concerns card acceptance.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4f6fce\": 0.0, \"candidate_931bf7\": 0.0, \"candidate_af633c\": 0.0, \"candidate_7ecd3c\": 1.0, \"candidate_29dbc8\": 0.0, \"candidate_11afa5\": 0.0, \"candidate_bce0e3\": 0.0, \"candidate_c2de1e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7755ea23057b640163391bf2:choice", "group_id": "fresh12:banking77:7755ea23057b640163391bf2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I'd rather not verify my identity.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5560a4\": \"The user's request concerns country support.\", \"candidate_420b6d\": \"The user's request concerns pin blocked.\", \"candidate_347d1e\": \"The user's request concerns unable to verify identity.\", \"candidate_4c2010\": \"The user's request concerns edit personal details.\", \"candidate_71328c\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_e6d3e4\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_55acc9\": \"The user's request concerns why verify identity.\", \"candidate_85b10c\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5560a4\": 0.0, \"candidate_420b6d\": 0.0, \"candidate_347d1e\": 0.0, \"candidate_4c2010\": 0.0, \"candidate_71328c\": 0.0, \"candidate_e6d3e4\": 0.0, \"candidate_55acc9\": 1.0, \"candidate_85b10c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:eb9fe948d720342dda06c7d1:choice", "group_id": "fresh12:banking77:eb9fe948d720342dda06c7d1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why did I need to verfiy my identity?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_feedea\": \"The user's request concerns transfer fee charged.\", \"candidate_8c27a4\": \"The user's request concerns country support.\", \"candidate_190ce0\": \"The user's request concerns declined transfer.\", \"candidate_f3d0c8\": \"The user's request concerns request refund.\", \"candidate_8e8193\": \"The user's request concerns extra charge on statement.\", \"candidate_d6fb24\": \"The user's request concerns why verify identity.\", \"candidate_0f87b8\": \"The user's request concerns declined card payment.\", \"candidate_06c1c0\": \"The user's request concerns pending transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_feedea\": 0.0, \"candidate_8c27a4\": 0.0, \"candidate_190ce0\": 0.0, \"candidate_f3d0c8\": 0.0, \"candidate_8e8193\": 0.0, \"candidate_d6fb24\": 1.0, \"candidate_0f87b8\": 0.0, \"candidate_06c1c0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:21036474c260a009252e2c56:choice", "group_id": "fresh12:banking77:21036474c260a009252e2c56", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why do you require all my identity details?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_86b46b\": \"The user's request concerns getting spare card.\", \"candidate_951ea9\": \"The user's request concerns card delivery estimate.\", \"candidate_1e6ab9\": \"The user's request concerns card arrival.\", \"candidate_ce1d4f\": \"The user's request concerns cash withdrawal charge.\", \"candidate_6c2357\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_b035bd\": \"The user's request concerns failed transfer.\", \"candidate_0f674a\": \"The user's request concerns card about to expire.\", \"candidate_0800a5\": \"The user's request concerns why verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_86b46b\": 0.0, \"candidate_951ea9\": 0.0, \"candidate_1e6ab9\": 0.0, \"candidate_ce1d4f\": 0.0, \"candidate_6c2357\": 0.0, \"candidate_b035bd\": 0.0, \"candidate_0f674a\": 0.0, \"candidate_0800a5\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e5707678b94af65a35600f3a:choice", "group_id": "fresh12:banking77:e5707678b94af65a35600f3a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Do I really need to verify my identity?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0322cc\": \"The user's request concerns failed transfer.\", \"candidate_13418e\": \"The user's request concerns wrong amount of cash received.\", \"candidate_ac1fa2\": \"The user's request concerns pending top up.\", \"candidate_3a6ba6\": \"The user's request concerns why verify identity.\", \"candidate_70e3da\": \"The user's request concerns lost or stolen card.\", \"candidate_3da696\": \"The user's request concerns automatic top up.\", \"candidate_43ffa5\": \"The user's request concerns card not working.\", \"candidate_277dca\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0322cc\": 0.0, \"candidate_13418e\": 0.0, \"candidate_ac1fa2\": 0.0, \"candidate_3a6ba6\": 1.0, \"candidate_70e3da\": 0.0, \"candidate_3da696\": 0.0, \"candidate_43ffa5\": 0.0, \"candidate_277dca\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b7dff2baf5b5d2e5b80093a0:choice", "group_id": "fresh12:banking77:b7dff2baf5b5d2e5b80093a0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"do I need to verify my identity every time I log into my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_995609\": \"The user's request concerns passcode forgotten.\", \"candidate_61eb6b\": \"The user's request concerns country support.\", \"candidate_42454e\": \"The user's request concerns compromised card.\", \"candidate_94a1b7\": \"The user's request concerns order physical card.\", \"candidate_4ded31\": \"The user's request concerns beneficiary not allowed.\", \"candidate_975d44\": \"The user's request concerns why verify identity.\", \"candidate_7869b4\": \"The user's request concerns activate my card.\", \"candidate_c97a56\": \"The user's request concerns terminate account.\"}}}}", "targets_json": "{\"decision\": {\"candidate_995609\": 0.0, \"candidate_61eb6b\": 0.0, \"candidate_42454e\": 0.0, \"candidate_94a1b7\": 0.0, \"candidate_4ded31\": 0.0, \"candidate_975d44\": 1.0, \"candidate_7869b4\": 0.0, \"candidate_c97a56\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a19a451f62aa87a963c066e4:choice", "group_id": "fresh12:banking77:a19a451f62aa87a963c066e4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My ID won't be verified!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bd6192\": \"The user's request concerns declined cash withdrawal.\", \"candidate_51f7c4\": \"The user's request concerns activate my card.\", \"candidate_c0b351\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_f809d2\": \"The user's request concerns verify source of funds.\", \"candidate_ff7027\": \"The user's request concerns receiving money.\", \"candidate_f2292b\": \"The user's request concerns pin blocked.\", \"candidate_fdba45\": \"The user's request concerns unable to verify identity.\", \"candidate_abbad0\": \"The user's request concerns card acceptance.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bd6192\": 0.0, \"candidate_51f7c4\": 0.0, \"candidate_c0b351\": 0.0, \"candidate_f809d2\": 0.0, \"candidate_ff7027\": 0.0, \"candidate_f2292b\": 0.0, \"candidate_fdba45\": 1.0, \"candidate_abbad0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0410ae66e0021c44970a2b5b:choice", "group_id": "fresh12:banking77:0410ae66e0021c44970a2b5b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why isn't my ID being verified?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7870a0\": \"The user's request concerns unable to verify identity.\", \"candidate_c6a748\": \"The user's request concerns exchange via app.\", \"candidate_237c43\": \"The user's request concerns topping up by card.\", \"candidate_e7495a\": \"The user's request concerns card payment not recognised.\", \"candidate_b5ce67\": \"The user's request concerns exchange charge.\", \"candidate_0c3c66\": \"The user's request concerns card delivery estimate.\", \"candidate_34e260\": \"The user's request concerns card acceptance.\", \"candidate_998b2a\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7870a0\": 1.0, \"candidate_c6a748\": 0.0, \"candidate_237c43\": 0.0, \"candidate_e7495a\": 0.0, \"candidate_b5ce67\": 0.0, \"candidate_0c3c66\": 0.0, \"candidate_34e260\": 0.0, \"candidate_998b2a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f8449ea8fb2c7834836d7453:choice", "group_id": "fresh12:banking77:f8449ea8fb2c7834836d7453", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I can't prove my identity.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6a6372\": \"The user's request concerns unable to verify identity.\", \"candidate_54b99f\": \"The user's request concerns pending transfer.\", \"candidate_0a000c\": \"The user's request concerns country support.\", \"candidate_4615d6\": \"The user's request concerns change pin.\", \"candidate_74b33f\": \"The user's request concerns lost or stolen phone.\", \"candidate_242565\": \"The user's request concerns activate my card.\", \"candidate_adb62d\": \"The user's request concerns why verify identity.\", \"candidate_7f0dc8\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6a6372\": 1.0, \"candidate_54b99f\": 0.0, \"candidate_0a000c\": 0.0, \"candidate_4615d6\": 0.0, \"candidate_74b33f\": 0.0, \"candidate_242565\": 0.0, \"candidate_adb62d\": 0.0, \"candidate_7f0dc8\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1aaf3e4bdeb2edb26a2ae89e:choice", "group_id": "fresh12:banking77:1aaf3e4bdeb2edb26a2ae89e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I cannot verify my identity\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1b6639\": \"The user's request concerns wrong amount of cash received.\", \"candidate_433bad\": \"The user's request concerns order physical card.\", \"candidate_0358ce\": \"The user's request concerns cash withdrawal charge.\", \"candidate_c28255\": \"The user's request concerns declined cash withdrawal.\", \"candidate_ced575\": \"The user's request concerns card about to expire.\", \"candidate_ed70cc\": \"The user's request concerns Refund not showing up.\", \"candidate_f61eea\": \"The user's request concerns unable to verify identity.\", \"candidate_85aa56\": \"The user's request concerns card payment wrong exchange rate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1b6639\": 0.0, \"candidate_433bad\": 0.0, \"candidate_0358ce\": 0.0, \"candidate_c28255\": 0.0, \"candidate_ced575\": 0.0, \"candidate_ed70cc\": 0.0, \"candidate_f61eea\": 1.0, \"candidate_85aa56\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0b80c281d752c0fc0519431f:choice", "group_id": "fresh12:banking77:0b80c281d752c0fc0519431f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I am having issues with identity verification\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_459ffa\": \"The user's request concerns cancel transfer.\", \"candidate_8ccf1c\": \"The user's request concerns card payment fee charged.\", \"candidate_2a920b\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_179812\": \"The user's request concerns declined transfer.\", \"candidate_7edc84\": \"The user's request concerns unable to verify identity.\", \"candidate_f174d1\": \"The user's request concerns declined card payment.\", \"candidate_3e4bf1\": \"The user's request concerns cash withdrawal charge.\", \"candidate_a50dba\": \"The user's request concerns exchange rate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_459ffa\": 0.0, \"candidate_8ccf1c\": 0.0, \"candidate_2a920b\": 0.0, \"candidate_179812\": 0.0, \"candidate_7edc84\": 1.0, \"candidate_f174d1\": 0.0, \"candidate_3e4bf1\": 0.0, \"candidate_a50dba\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:33d687e7c4d3c75931ef4cb3:choice", "group_id": "fresh12:banking77:33d687e7c4d3c75931ef4cb3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"app doesnt think its me\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4e4669\": \"The user's request concerns compromised card.\", \"candidate_64ccc6\": \"The user's request concerns unable to verify identity.\", \"candidate_a9b50a\": \"The user's request concerns lost or stolen card.\", \"candidate_eed438\": \"The user's request concerns exchange via app.\", \"candidate_49250b\": \"The user's request concerns get physical card.\", \"candidate_4f38ed\": \"The user's request concerns card acceptance.\", \"candidate_74ba0a\": \"The user's request concerns edit personal details.\", \"candidate_c406d0\": \"The user's request concerns cash withdrawal not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4e4669\": 0.0, \"candidate_64ccc6\": 1.0, \"candidate_a9b50a\": 0.0, \"candidate_eed438\": 0.0, \"candidate_49250b\": 0.0, \"candidate_4f38ed\": 0.0, \"candidate_74ba0a\": 0.0, \"candidate_c406d0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f894751e136e37a95baf049f:choice", "group_id": "fresh12:banking77:f894751e136e37a95baf049f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"If I need a PIN for my card, where is it located?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c9ea2a\": \"The user's request concerns supported cards and currencies.\", \"candidate_8201aa\": \"The user's request concerns pending cash withdrawal.\", \"candidate_c2bcfe\": \"The user's request concerns why verify identity.\", \"candidate_5ae6e7\": \"The user's request concerns declined transfer.\", \"candidate_1c9483\": \"The user's request concerns compromised card.\", \"candidate_40393f\": \"The user's request concerns getting spare card.\", \"candidate_81fe5c\": \"The user's request concerns get physical card.\", \"candidate_828e02\": \"The user's request concerns balance not updated after bank transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c9ea2a\": 0.0, \"candidate_8201aa\": 0.0, \"candidate_c2bcfe\": 0.0, \"candidate_5ae6e7\": 0.0, \"candidate_1c9483\": 0.0, \"candidate_40393f\": 0.0, \"candidate_81fe5c\": 1.0, \"candidate_828e02\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:937f476c62546d1e01d98034:choice", "group_id": "fresh12:banking77:937f476c62546d1e01d98034", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Are PIN separately?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_711769\": \"The user's request concerns cancel transfer.\", \"candidate_5f7594\": \"The user's request concerns get physical card.\", \"candidate_1e2443\": \"The user's request concerns card delivery estimate.\", \"candidate_1f4ea1\": \"The user's request concerns age limit.\", \"candidate_a52c2f\": \"The user's request concerns pending cash withdrawal.\", \"candidate_9227eb\": \"The user's request concerns card swallowed.\", \"candidate_a714ec\": \"The user's request concerns card about to expire.\", \"candidate_aa67d5\": \"The user's request concerns reverted card payment?.\"}}}}", "targets_json": "{\"decision\": {\"candidate_711769\": 0.0, \"candidate_5f7594\": 1.0, \"candidate_1e2443\": 0.0, \"candidate_1f4ea1\": 0.0, \"candidate_a52c2f\": 0.0, \"candidate_9227eb\": 0.0, \"candidate_a714ec\": 0.0, \"candidate_aa67d5\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1325afad1ccf5c10388e80f8:choice", "group_id": "fresh12:banking77:1325afad1ccf5c10388e80f8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"where can user find pin?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7afca2\": \"The user's request concerns get physical card.\", \"candidate_0969db\": \"The user's request concerns fiat currency support.\", \"candidate_440bce\": \"The user's request concerns pending cash withdrawal.\", \"candidate_c3a480\": \"The user's request concerns lost or stolen card.\", \"candidate_a0fb0b\": \"The user's request concerns card about to expire.\", \"candidate_081cfb\": \"The user's request concerns transfer fee charged.\", \"candidate_5cc284\": \"The user's request concerns order physical card.\", \"candidate_9b7657\": \"The user's request concerns cash withdrawal not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7afca2\": 1.0, \"candidate_0969db\": 0.0, \"candidate_440bce\": 0.0, \"candidate_c3a480\": 0.0, \"candidate_a0fb0b\": 0.0, \"candidate_081cfb\": 0.0, \"candidate_5cc284\": 0.0, \"candidate_9b7657\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8e310424cf9c3f191afede42:choice", "group_id": "fresh12:banking77:8e310424cf9c3f191afede42", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is the PIN delivered separately?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e0c000\": \"The user's request concerns pending transfer.\", \"candidate_36c770\": \"The user's request concerns compromised card.\", \"candidate_546336\": \"The user's request concerns receiving money.\", \"candidate_b99a72\": \"The user's request concerns card swallowed.\", \"candidate_112340\": \"The user's request concerns top up failed.\", \"candidate_a862ab\": \"The user's request concerns get physical card.\", \"candidate_383ee3\": \"The user's request concerns country support.\", \"candidate_be42ee\": \"The user's request concerns transfer timing.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e0c000\": 0.0, \"candidate_36c770\": 0.0, \"candidate_546336\": 0.0, \"candidate_b99a72\": 0.0, \"candidate_112340\": 0.0, \"candidate_a862ab\": 1.0, \"candidate_383ee3\": 0.0, \"candidate_be42ee\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:93e4da07e164f5352712596a:choice", "group_id": "fresh12:banking77:93e4da07e164f5352712596a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is my PIN located on my account somewhere?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_42130b\": \"The user's request concerns change pin.\", \"candidate_a9ddec\": \"The user's request concerns get physical card.\", \"candidate_a5d31d\": \"The user's request concerns why verify identity.\", \"candidate_54ef29\": \"The user's request concerns get disposable virtual card.\", \"candidate_13ab60\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_8b9496\": \"The user's request concerns extra charge on statement.\", \"candidate_289d9b\": \"The user's request concerns unable to verify identity.\", \"candidate_a9bada\": \"The user's request concerns transfer into account.\"}}}}", "targets_json": "{\"decision\": {\"candidate_42130b\": 0.0, \"candidate_a9ddec\": 1.0, \"candidate_a5d31d\": 0.0, \"candidate_54ef29\": 0.0, \"candidate_13ab60\": 0.0, \"candidate_8b9496\": 0.0, \"candidate_289d9b\": 0.0, \"candidate_a9bada\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:fc19f59a65d9a8b930316bef:choice", "group_id": "fresh12:banking77:fc19f59a65d9a8b930316bef", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where can I find the card PIN?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c059ef\": \"The user's request concerns atm support.\", \"candidate_204a03\": \"The user's request concerns top up failed.\", \"candidate_61a74b\": \"The user's request concerns virtual card not working.\", \"candidate_2259b8\": \"The user's request concerns transfer not received by recipient.\", \"candidate_50ea2a\": \"The user's request concerns transfer timing.\", \"candidate_a71725\": \"The user's request concerns get physical card.\", \"candidate_5c17d1\": \"The user's request concerns lost or stolen phone.\", \"candidate_e7f0cf\": \"The user's request concerns card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c059ef\": 0.0, \"candidate_204a03\": 0.0, \"candidate_61a74b\": 0.0, \"candidate_2259b8\": 0.0, \"candidate_50ea2a\": 0.0, \"candidate_a71725\": 1.0, \"candidate_5c17d1\": 0.0, \"candidate_e7f0cf\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4758968ff3b0b3a7e9d94341:choice", "group_id": "fresh12:banking77:4758968ff3b0b3a7e9d94341", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Do you use Mastercard or Visa?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8fc8a7\": \"The user's request concerns visa or mastercard.\", \"candidate_1ed1a0\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_4697d5\": \"The user's request concerns card not working.\", \"candidate_3886fd\": \"The user's request concerns country support.\", \"candidate_f93ac6\": \"The user's request concerns wrong amount of cash received.\", \"candidate_6b3642\": \"The user's request concerns terminate account.\", \"candidate_45ad58\": \"The user's request concerns disposable card limits.\", \"candidate_9a1caf\": \"The user's request concerns top up by cash or cheque.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8fc8a7\": 1.0, \"candidate_1ed1a0\": 0.0, \"candidate_4697d5\": 0.0, \"candidate_3886fd\": 0.0, \"candidate_f93ac6\": 0.0, \"candidate_6b3642\": 0.0, \"candidate_45ad58\": 0.0, \"candidate_9a1caf\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:04b086bbade8d54509e9ef3d:choice", "group_id": "fresh12:banking77:04b086bbade8d54509e9ef3d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What major card payments are accepted?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0acfba\": \"The user's request concerns fiat currency support.\", \"candidate_f0a971\": \"The user's request concerns Refund not showing up.\", \"candidate_362595\": \"The user's request concerns cash withdrawal charge.\", \"candidate_a4003e\": \"The user's request concerns transfer fee charged.\", \"candidate_e773d6\": \"The user's request concerns visa or mastercard.\", \"candidate_79e3e9\": \"The user's request concerns transfer not received by recipient.\", \"candidate_831492\": \"The user's request concerns declined transfer.\", \"candidate_86f87a\": \"The user's request concerns unable to verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0acfba\": 0.0, \"candidate_f0a971\": 0.0, \"candidate_362595\": 0.0, \"candidate_a4003e\": 0.0, \"candidate_e773d6\": 1.0, \"candidate_79e3e9\": 0.0, \"candidate_831492\": 0.0, \"candidate_86f87a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c2b0a4ae1caf4fb5ea6d8962:choice", "group_id": "fresh12:banking77:c2b0a4ae1caf4fb5ea6d8962", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Please give me both a visa and a master card.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2148c1\": \"The user's request concerns fiat currency support.\", \"candidate_79a9d2\": \"The user's request concerns lost or stolen card.\", \"candidate_737891\": \"The user's request concerns disposable card limits.\", \"candidate_d66e4e\": \"The user's request concerns verify my identity.\", \"candidate_ef3121\": \"The user's request concerns top up failed.\", \"candidate_b84c42\": \"The user's request concerns getting spare card.\", \"candidate_75f83a\": \"The user's request concerns visa or mastercard.\", \"candidate_0ffca4\": \"The user's request concerns verify top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2148c1\": 0.0, \"candidate_79a9d2\": 0.0, \"candidate_737891\": 0.0, \"candidate_d66e4e\": 0.0, \"candidate_ef3121\": 0.0, \"candidate_b84c42\": 0.0, \"candidate_75f83a\": 1.0, \"candidate_0ffca4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1410f40843b42129cc10e98e:choice", "group_id": "fresh12:banking77:1410f40843b42129cc10e98e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is it possible to get a Visa here?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_919a71\": \"The user's request concerns age limit.\", \"candidate_d226ff\": \"The user's request concerns pending cash withdrawal.\", \"candidate_7c11de\": \"The user's request concerns lost or stolen phone.\", \"candidate_a3a0d6\": \"The user's request concerns verify top up.\", \"candidate_274180\": \"The user's request concerns visa or mastercard.\", \"candidate_1a0f82\": \"The user's request concerns pin blocked.\", \"candidate_bedce6\": \"The user's request concerns disposable card limits.\", \"candidate_72b6ec\": \"The user's request concerns automatic top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_919a71\": 0.0, \"candidate_d226ff\": 0.0, \"candidate_7c11de\": 0.0, \"candidate_a3a0d6\": 0.0, \"candidate_274180\": 1.0, \"candidate_1a0f82\": 0.0, \"candidate_bedce6\": 0.0, \"candidate_72b6ec\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:771b9f14d62967b6ea5c3a55:choice", "group_id": "fresh12:banking77:771b9f14d62967b6ea5c3a55", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I would prefer a visa card\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fc5132\": \"The user's request concerns card delivery estimate.\", \"candidate_96cc8e\": \"The user's request concerns declined transfer.\", \"candidate_158d04\": \"The user's request concerns passcode forgotten.\", \"candidate_c64bda\": \"The user's request concerns transfer fee charged.\", \"candidate_5f4da7\": \"The user's request concerns receiving money.\", \"candidate_aa8dca\": \"The user's request concerns visa or mastercard.\", \"candidate_a9a45b\": \"The user's request concerns beneficiary not allowed.\", \"candidate_a561bd\": \"The user's request concerns edit personal details.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fc5132\": 0.0, \"candidate_96cc8e\": 0.0, \"candidate_158d04\": 0.0, \"candidate_c64bda\": 0.0, \"candidate_5f4da7\": 0.0, \"candidate_aa8dca\": 1.0, \"candidate_a9a45b\": 0.0, \"candidate_a561bd\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:15da714c094bb48b1c40e342:choice", "group_id": "fresh12:banking77:15da714c094bb48b1c40e342", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What steps must I follow to get a Visa credit card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2e894a\": \"The user's request concerns visa or mastercard.\", \"candidate_c7627a\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_d7b9fe\": \"The user's request concerns getting virtual card.\", \"candidate_75d9de\": \"The user's request concerns compromised card.\", \"candidate_bb6af0\": \"The user's request concerns wrong amount of cash received.\", \"candidate_2e824c\": \"The user's request concerns unable to verify identity.\", \"candidate_97ec82\": \"The user's request concerns declined card payment.\", \"candidate_1c37c4\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2e894a\": 1.0, \"candidate_c7627a\": 0.0, \"candidate_d7b9fe\": 0.0, \"candidate_75d9de\": 0.0, \"candidate_bb6af0\": 0.0, \"candidate_2e824c\": 0.0, \"candidate_97ec82\": 0.0, \"candidate_1c37c4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3a5e009447642d3b6ea5d2ae:choice", "group_id": "fresh12:banking77:3a5e009447642d3b6ea5d2ae", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I give my friends access to my account so they can top it up for me?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_259f3e\": \"The user's request concerns age limit.\", \"candidate_ba73b0\": \"The user's request concerns card payment fee charged.\", \"candidate_34b32a\": \"The user's request concerns disposable card limits.\", \"candidate_5eedb5\": \"The user's request concerns card not working.\", \"candidate_eaeb80\": \"The user's request concerns topping up by card.\", \"candidate_598921\": \"The user's request concerns country support.\", \"candidate_eccddf\": \"The user's request concerns contactless not working.\", \"candidate_6e5124\": \"The user's request concerns supported cards and currencies.\"}}}}", "targets_json": "{\"decision\": {\"candidate_259f3e\": 0.0, \"candidate_ba73b0\": 0.0, \"candidate_34b32a\": 0.0, \"candidate_5eedb5\": 0.0, \"candidate_eaeb80\": 1.0, \"candidate_598921\": 0.0, \"candidate_eccddf\": 0.0, \"candidate_6e5124\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8ebebc1ae5e11b0cc9f73fe2:choice", "group_id": "fresh12:banking77:8ebebc1ae5e11b0cc9f73fe2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where's the money that got charged to my card? It's not showing up in my account balance\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e1bde3\": \"The user's request concerns card about to expire.\", \"candidate_81c895\": \"The user's request concerns topping up by card.\", \"candidate_eaba06\": \"The user's request concerns order physical card.\", \"candidate_788de6\": \"The user's request concerns cash withdrawal charge.\", \"candidate_5e7877\": \"The user's request concerns verify source of funds.\", \"candidate_75d717\": \"The user's request concerns transfer into account.\", \"candidate_d5f9bb\": \"The user's request concerns supported cards and currencies.\", \"candidate_97f3aa\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e1bde3\": 0.0, \"candidate_81c895\": 1.0, \"candidate_eaba06\": 0.0, \"candidate_788de6\": 0.0, \"candidate_5e7877\": 0.0, \"candidate_75d717\": 0.0, \"candidate_d5f9bb\": 0.0, \"candidate_97f3aa\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0999f47f31dd6aeaf85e6181:choice", "group_id": "fresh12:banking77:0999f47f31dd6aeaf85e6181", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I topped up my card but the money disappeared.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_be78fe\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_8884d0\": \"The user's request concerns contactless not working.\", \"candidate_af10dc\": \"The user's request concerns topping up by card.\", \"candidate_8e8b41\": \"The user's request concerns top up by card charge.\", \"candidate_90cec9\": \"The user's request concerns top up limits.\", \"candidate_30a645\": \"The user's request concerns change pin.\", \"candidate_4c2c2e\": \"The user's request concerns exchange rate.\", \"candidate_46cb60\": \"The user's request concerns top up failed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_be78fe\": 0.0, \"candidate_8884d0\": 0.0, \"candidate_af10dc\": 1.0, \"candidate_8e8b41\": 0.0, \"candidate_90cec9\": 0.0, \"candidate_30a645\": 0.0, \"candidate_4c2c2e\": 0.0, \"candidate_46cb60\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f7aa2dcb4f75a6a5a6e6b4a3:choice", "group_id": "fresh12:banking77:f7aa2dcb4f75a6a5a6e6b4a3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can I top up by card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d5deff\": \"The user's request concerns topping up by card.\", \"candidate_1bee19\": \"The user's request concerns get disposable virtual card.\", \"candidate_c22545\": \"The user's request concerns passcode forgotten.\", \"candidate_558e84\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_8d4010\": \"The user's request concerns terminate account.\", \"candidate_a0e53e\": \"The user's request concerns edit personal details.\", \"candidate_c49e51\": \"The user's request concerns wrong amount of cash received.\", \"candidate_5dfe4c\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d5deff\": 1.0, \"candidate_1bee19\": 0.0, \"candidate_c22545\": 0.0, \"candidate_558e84\": 0.0, \"candidate_8d4010\": 0.0, \"candidate_a0e53e\": 0.0, \"candidate_c49e51\": 0.0, \"candidate_5dfe4c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ca449f6978dc47115593bf40:choice", "group_id": "fresh12:banking77:ca449f6978dc47115593bf40", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I top up my card using your app? I'm new to this.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5a6eb8\": \"The user's request concerns supported cards and currencies.\", \"candidate_3db453\": \"The user's request concerns verify my identity.\", \"candidate_23762f\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_fb3baf\": \"The user's request concerns card about to expire.\", \"candidate_789335\": \"The user's request concerns activate my card.\", \"candidate_a345f4\": \"The user's request concerns topping up by card.\", \"candidate_30da61\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_22cfad\": \"The user's request concerns virtual card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5a6eb8\": 0.0, \"candidate_3db453\": 0.0, \"candidate_23762f\": 0.0, \"candidate_fb3baf\": 0.0, \"candidate_789335\": 0.0, \"candidate_a345f4\": 1.0, \"candidate_30da61\": 0.0, \"candidate_22cfad\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e2416a57bad824e71d6db97a:choice", "group_id": "fresh12:banking77:e2416a57bad824e71d6db97a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Use credit card to transfer money\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_de95b2\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_51c488\": \"The user's request concerns beneficiary not allowed.\", \"candidate_a0ba37\": \"The user's request concerns reverted card payment?.\", \"candidate_81b77a\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_2a628f\": \"The user's request concerns topping up by card.\", \"candidate_e79094\": \"The user's request concerns getting virtual card.\", \"candidate_e3fc98\": \"The user's request concerns exchange rate.\", \"candidate_b1f055\": \"The user's request concerns unable to verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_de95b2\": 0.0, \"candidate_51c488\": 0.0, \"candidate_a0ba37\": 0.0, \"candidate_81b77a\": 0.0, \"candidate_2a628f\": 1.0, \"candidate_e79094\": 0.0, \"candidate_e3fc98\": 0.0, \"candidate_b1f055\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:85e73fa0df99d1c47dc7bb16:choice", "group_id": "fresh12:banking77:85e73fa0df99d1c47dc7bb16", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How many transactions can I do with one disposable card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9a3534\": \"The user's request concerns disposable card limits.\", \"candidate_684c19\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_ac709a\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_ffdbe1\": \"The user's request concerns card acceptance.\", \"candidate_bd6dad\": \"The user's request concerns change pin.\", \"candidate_cafa77\": \"The user's request concerns request refund.\", \"candidate_883565\": \"The user's request concerns card about to expire.\", \"candidate_57e1d5\": \"The user's request concerns transfer into account.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9a3534\": 1.0, \"candidate_684c19\": 0.0, \"candidate_ac709a\": 0.0, \"candidate_ffdbe1\": 0.0, \"candidate_bd6dad\": 0.0, \"candidate_cafa77\": 0.0, \"candidate_883565\": 0.0, \"candidate_57e1d5\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c1bed0cec083e8a9760a595f:choice", "group_id": "fresh12:banking77:c1bed0cec083e8a9760a595f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What if I need multiple disposable cards?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5257d6\": \"The user's request concerns card about to expire.\", \"candidate_993230\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_ca7895\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_74802d\": \"The user's request concerns disposable card limits.\", \"candidate_91ecf2\": \"The user's request concerns verify my identity.\", \"candidate_47a8fd\": \"The user's request concerns apple pay or google pay.\", \"candidate_c03884\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_e549a3\": \"The user's request concerns transfer not received by recipient.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5257d6\": 0.0, \"candidate_993230\": 0.0, \"candidate_ca7895\": 0.0, \"candidate_74802d\": 1.0, \"candidate_91ecf2\": 0.0, \"candidate_47a8fd\": 0.0, \"candidate_c03884\": 0.0, \"candidate_e549a3\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:aeaf5e61361ab9a49f8541e8:choice", "group_id": "fresh12:banking77:aeaf5e61361ab9a49f8541e8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need to use my disposable card multiple times a day, is there a cutoff limit?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c6d0f6\": \"The user's request concerns exchange charge.\", \"candidate_0c72f6\": \"The user's request concerns exchange rate.\", \"candidate_ba51fc\": \"The user's request concerns disposable card limits.\", \"candidate_041bee\": \"The user's request concerns get disposable virtual card.\", \"candidate_b50b41\": \"The user's request concerns top up limits.\", \"candidate_6a4427\": \"The user's request concerns transfer timing.\", \"candidate_f66808\": \"The user's request concerns request refund.\", \"candidate_c4540a\": \"The user's request concerns transfer fee charged.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c6d0f6\": 0.0, \"candidate_0c72f6\": 0.0, \"candidate_ba51fc\": 1.0, \"candidate_041bee\": 0.0, \"candidate_b50b41\": 0.0, \"candidate_6a4427\": 0.0, \"candidate_f66808\": 0.0, \"candidate_c4540a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a0261965c02c5f9bd225ced7:choice", "group_id": "fresh12:banking77:a0261965c02c5f9bd225ced7", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need to create several disposable cards per day.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_97961f\": \"The user's request concerns unable to verify identity.\", \"candidate_865582\": \"The user's request concerns verify my identity.\", \"candidate_fee4d0\": \"The user's request concerns age limit.\", \"candidate_602ce3\": \"The user's request concerns change pin.\", \"candidate_5dafd2\": \"The user's request concerns visa or mastercard.\", \"candidate_b45b2a\": \"The user's request concerns card acceptance.\", \"candidate_1332d8\": \"The user's request concerns disposable card limits.\", \"candidate_54db30\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_97961f\": 0.0, \"candidate_865582\": 0.0, \"candidate_fee4d0\": 0.0, \"candidate_602ce3\": 0.0, \"candidate_5dafd2\": 0.0, \"candidate_b45b2a\": 0.0, \"candidate_1332d8\": 1.0, \"candidate_54db30\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8d26827ab94429b0c465a7b8:choice", "group_id": "fresh12:banking77:8d26827ab94429b0c465a7b8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How many disposable cards per day am I allowed to have?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c82851\": \"The user's request concerns exchange via app.\", \"candidate_293cdc\": \"The user's request concerns disposable card limits.\", \"candidate_813f51\": \"The user's request concerns virtual card not working.\", \"candidate_213839\": \"The user's request concerns terminate account.\", \"candidate_d72815\": \"The user's request concerns why verify identity.\", \"candidate_6ccaf0\": \"The user's request concerns transaction charged twice.\", \"candidate_bc84a1\": \"The user's request concerns verify source of funds.\", \"candidate_466444\": \"The user's request concerns cash withdrawal not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c82851\": 0.0, \"candidate_293cdc\": 1.0, \"candidate_813f51\": 0.0, \"candidate_213839\": 0.0, \"candidate_d72815\": 0.0, \"candidate_6ccaf0\": 0.0, \"candidate_bc84a1\": 0.0, \"candidate_466444\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:75f7dd58850cf711d318c331:choice", "group_id": "fresh12:banking77:75f7dd58850cf711d318c331", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a limit to how many transactions I can make on one disposable card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_44e7fd\": \"The user's request concerns exchange rate.\", \"candidate_6825a9\": \"The user's request concerns top up reverted.\", \"candidate_548ab2\": \"The user's request concerns top up limits.\", \"candidate_e6927a\": \"The user's request concerns terminate account.\", \"candidate_fce672\": \"The user's request concerns reverted card payment?.\", \"candidate_520453\": \"The user's request concerns pending top up.\", \"candidate_8372de\": \"The user's request concerns lost or stolen phone.\", \"candidate_e7e0c8\": \"The user's request concerns disposable card limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_44e7fd\": 0.0, \"candidate_6825a9\": 0.0, \"candidate_548ab2\": 0.0, \"candidate_e6927a\": 0.0, \"candidate_fce672\": 0.0, \"candidate_520453\": 0.0, \"candidate_8372de\": 0.0, \"candidate_e7e0c8\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0e6c39079908fca07c81bf5e:choice", "group_id": "fresh12:banking77:0e6c39079908fca07c81bf5e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"someone is using my account\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_57c2a5\": \"The user's request concerns compromised card.\", \"candidate_8ef8f6\": \"The user's request concerns verify my identity.\", \"candidate_721550\": \"The user's request concerns declined transfer.\", \"candidate_5f488f\": \"The user's request concerns card delivery estimate.\", \"candidate_a60a0c\": \"The user's request concerns transfer timing.\", \"candidate_05ffaf\": \"The user's request concerns edit personal details.\", \"candidate_5673ab\": \"The user's request concerns lost or stolen phone.\", \"candidate_4b59e1\": \"The user's request concerns cash withdrawal charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_57c2a5\": 1.0, \"candidate_8ef8f6\": 0.0, \"candidate_721550\": 0.0, \"candidate_5f488f\": 0.0, \"candidate_a60a0c\": 0.0, \"candidate_05ffaf\": 0.0, \"candidate_5673ab\": 0.0, \"candidate_4b59e1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:33a5c8494f870f7ed3213ebe:choice", "group_id": "fresh12:banking77:33a5c8494f870f7ed3213ebe", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card is being used online to make a bunch of purchases! Someone else is using it fraudulently. Can you help me freeze it?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_520ef4\": \"The user's request concerns activate my card.\", \"candidate_5a1e27\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_a06417\": \"The user's request concerns top up limits.\", \"candidate_47b969\": \"The user's request concerns card not working.\", \"candidate_41c8bf\": \"The user's request concerns card swallowed.\", \"candidate_de9070\": \"The user's request concerns country support.\", \"candidate_3b0b86\": \"The user's request concerns lost or stolen card.\", \"candidate_23544d\": \"The user's request concerns compromised card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_520ef4\": 0.0, \"candidate_5a1e27\": 0.0, \"candidate_a06417\": 0.0, \"candidate_47b969\": 0.0, \"candidate_41c8bf\": 0.0, \"candidate_de9070\": 0.0, \"candidate_3b0b86\": 0.0, \"candidate_23544d\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:319dcbe7a9945381e7d9b5db:choice", "group_id": "fresh12:banking77:319dcbe7a9945381e7d9b5db", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Someone else may be using my card.  There are transactions I don't recognize.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a91f1d\": \"The user's request concerns exchange charge.\", \"candidate_193386\": \"The user's request concerns top up by cash or cheque.\", \"candidate_2c7e08\": \"The user's request concerns verify my identity.\", \"candidate_72e9da\": \"The user's request concerns pending cash withdrawal.\", \"candidate_d03363\": \"The user's request concerns card payment not recognised.\", \"candidate_afc991\": \"The user's request concerns pending top up.\", \"candidate_3751d3\": \"The user's request concerns compromised card.\", \"candidate_e01781\": \"The user's request concerns wrong amount of cash received.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a91f1d\": 0.0, \"candidate_193386\": 0.0, \"candidate_2c7e08\": 0.0, \"candidate_72e9da\": 0.0, \"candidate_d03363\": 0.0, \"candidate_afc991\": 0.0, \"candidate_3751d3\": 1.0, \"candidate_e01781\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c0746ee1648be657b1766423:choice", "group_id": "fresh12:banking77:c0746ee1648be657b1766423", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What do I do to stop unauthorized transactions on my card? I was never in the place from which the transactions on my bill were made and I never made them. Please help!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bbb70d\": \"The user's request concerns top up failed.\", \"candidate_363d01\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_6994bd\": \"The user's request concerns country support.\", \"candidate_42b779\": \"The user's request concerns card swallowed.\", \"candidate_140aef\": \"The user's request concerns cancel transfer.\", \"candidate_57ebc5\": \"The user's request concerns card payment fee charged.\", \"candidate_076361\": \"The user's request concerns compromised card.\", \"candidate_45c10a\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bbb70d\": 0.0, \"candidate_363d01\": 0.0, \"candidate_6994bd\": 0.0, \"candidate_42b779\": 0.0, \"candidate_140aef\": 0.0, \"candidate_57ebc5\": 0.0, \"candidate_076361\": 1.0, \"candidate_45c10a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2cc221f88309375c0d543cc0:choice", "group_id": "fresh12:banking77:2cc221f88309375c0d543cc0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I think someone has access to my card numbers that shouldn't.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_163db5\": \"The user's request concerns activate my card.\", \"candidate_0bce69\": \"The user's request concerns transaction charged twice.\", \"candidate_356060\": \"The user's request concerns lost or stolen card.\", \"candidate_e2ab8a\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_353589\": \"The user's request concerns card arrival.\", \"candidate_776f9a\": \"The user's request concerns pin blocked.\", \"candidate_ed684a\": \"The user's request concerns card swallowed.\", \"candidate_2bb72f\": \"The user's request concerns compromised card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_163db5\": 0.0, \"candidate_0bce69\": 0.0, \"candidate_356060\": 0.0, \"candidate_e2ab8a\": 0.0, \"candidate_353589\": 0.0, \"candidate_776f9a\": 0.0, \"candidate_ed684a\": 0.0, \"candidate_2bb72f\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:fd1abdaff966d9a4658a1e01:choice", "group_id": "fresh12:banking77:fd1abdaff966d9a4658a1e01", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There are transactions I didn't make, someone else must have used my card.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0234a4\": \"The user's request concerns declined transfer.\", \"candidate_51c6d2\": \"The user's request concerns compromised card.\", \"candidate_4f2fbd\": \"The user's request concerns card payment fee charged.\", \"candidate_a34c15\": \"The user's request concerns visa or mastercard.\", \"candidate_cd011e\": \"The user's request concerns beneficiary not allowed.\", \"candidate_e22c94\": \"The user's request concerns age limit.\", \"candidate_a6d8ab\": \"The user's request concerns pending card payment.\", \"candidate_470eef\": \"The user's request concerns Refund not showing up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0234a4\": 0.0, \"candidate_51c6d2\": 1.0, \"candidate_4f2fbd\": 0.0, \"candidate_a34c15\": 0.0, \"candidate_cd011e\": 0.0, \"candidate_e22c94\": 0.0, \"candidate_a6d8ab\": 0.0, \"candidate_470eef\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:db99393ebbc442944d8d8a28:choice", "group_id": "fresh12:banking77:db99393ebbc442944d8d8a28", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Which ATMs accept this card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8bde91\": \"The user's request concerns declined cash withdrawal.\", \"candidate_ae9abe\": \"The user's request concerns card arrival.\", \"candidate_41fb31\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_216cf8\": \"The user's request concerns activate my card.\", \"candidate_915675\": \"The user's request concerns lost or stolen phone.\", \"candidate_7fe4ac\": \"The user's request concerns beneficiary not allowed.\", \"candidate_66319b\": \"The user's request concerns atm support.\", \"candidate_d0d825\": \"The user's request concerns receiving money.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8bde91\": 0.0, \"candidate_ae9abe\": 0.0, \"candidate_41fb31\": 0.0, \"candidate_216cf8\": 0.0, \"candidate_915675\": 0.0, \"candidate_7fe4ac\": 0.0, \"candidate_66319b\": 1.0, \"candidate_d0d825\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a1a87ad26b81934f983f1543:choice", "group_id": "fresh12:banking77:a1a87ad26b81934f983f1543", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a list of ATM machines?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_122183\": \"The user's request concerns terminate account.\", \"candidate_ad8609\": \"The user's request concerns atm support.\", \"candidate_48faf6\": \"The user's request concerns wrong amount of cash received.\", \"candidate_85f69b\": \"The user's request concerns supported cards and currencies.\", \"candidate_22d88f\": \"The user's request concerns contactless not working.\", \"candidate_2f3180\": \"The user's request concerns compromised card.\", \"candidate_938a34\": \"The user's request concerns receiving money.\", \"candidate_948cab\": \"The user's request concerns change pin.\"}}}}", "targets_json": "{\"decision\": {\"candidate_122183\": 0.0, \"candidate_ad8609\": 1.0, \"candidate_48faf6\": 0.0, \"candidate_85f69b\": 0.0, \"candidate_22d88f\": 0.0, \"candidate_2f3180\": 0.0, \"candidate_938a34\": 0.0, \"candidate_948cab\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a560d4d5c15876c4ab2c96ff:choice", "group_id": "fresh12:banking77:a560d4d5c15876c4ab2c96ff", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where is the nearest bank machine?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4ab7e7\": \"The user's request concerns topping up by card.\", \"candidate_25742c\": \"The user's request concerns get physical card.\", \"candidate_305129\": \"The user's request concerns pending cash withdrawal.\", \"candidate_7dc12f\": \"The user's request concerns receiving money.\", \"candidate_32baa5\": \"The user's request concerns exchange charge.\", \"candidate_de1864\": \"The user's request concerns atm support.\", \"candidate_160cb8\": \"The user's request concerns country support.\", \"candidate_c45924\": \"The user's request concerns get disposable virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4ab7e7\": 0.0, \"candidate_25742c\": 0.0, \"candidate_305129\": 0.0, \"candidate_7dc12f\": 0.0, \"candidate_32baa5\": 0.0, \"candidate_de1864\": 1.0, \"candidate_160cb8\": 0.0, \"candidate_c45924\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2e31ef191c393b55f3745ba6:choice", "group_id": "fresh12:banking77:2e31ef191c393b55f3745ba6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there places where I can't withdraw money?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_deec12\": \"The user's request concerns verify top up.\", \"candidate_5a2689\": \"The user's request concerns change pin.\", \"candidate_09cb49\": \"The user's request concerns declined card payment.\", \"candidate_191121\": \"The user's request concerns get physical card.\", \"candidate_3b5fe4\": \"The user's request concerns activate my card.\", \"candidate_f5b040\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_0854ab\": \"The user's request concerns atm support.\", \"candidate_73d958\": \"The user's request concerns exchange via app.\"}}}}", "targets_json": "{\"decision\": {\"candidate_deec12\": 0.0, \"candidate_5a2689\": 0.0, \"candidate_09cb49\": 0.0, \"candidate_191121\": 0.0, \"candidate_3b5fe4\": 0.0, \"candidate_f5b040\": 0.0, \"candidate_0854ab\": 1.0, \"candidate_73d958\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:975015ae41573459fca5bfbb:choice", "group_id": "fresh12:banking77:975015ae41573459fca5bfbb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where is the closest Mastercard ATM?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0b36e1\": \"The user's request concerns automatic top up.\", \"candidate_86bafd\": \"The user's request concerns transfer fee charged.\", \"candidate_7c7ac0\": \"The user's request concerns transfer into account.\", \"candidate_13e6cc\": \"The user's request concerns disposable card limits.\", \"candidate_5db9f2\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_c2c270\": \"The user's request concerns atm support.\", \"candidate_c8984c\": \"The user's request concerns change pin.\", \"candidate_86a602\": \"The user's request concerns declined cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0b36e1\": 0.0, \"candidate_86bafd\": 0.0, \"candidate_7c7ac0\": 0.0, \"candidate_13e6cc\": 0.0, \"candidate_5db9f2\": 0.0, \"candidate_c2c270\": 1.0, \"candidate_c8984c\": 0.0, \"candidate_86a602\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2f987fb8f1347f1a839c27e8:choice", "group_id": "fresh12:banking77:2f987fb8f1347f1a839c27e8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"If I have this card, which ATMs can I go to?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2496eb\": \"The user's request concerns extra charge on statement.\", \"candidate_f5fc12\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_e31716\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_6185bc\": \"The user's request concerns order physical card.\", \"candidate_b9a7b4\": \"The user's request concerns terminate account.\", \"candidate_89db4c\": \"The user's request concerns pending top up.\", \"candidate_bedf81\": \"The user's request concerns lost or stolen phone.\", \"candidate_17b2b9\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2496eb\": 0.0, \"candidate_f5fc12\": 0.0, \"candidate_e31716\": 0.0, \"candidate_6185bc\": 0.0, \"candidate_b9a7b4\": 0.0, \"candidate_89db4c\": 0.0, \"candidate_bedf81\": 0.0, \"candidate_17b2b9\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5ab3be3591ffdd73bafcac15:choice", "group_id": "fresh12:banking77:5ab3be3591ffdd73bafcac15", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Unknown direct deposit\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9612c9\": \"The user's request concerns visa or mastercard.\", \"candidate_0f7292\": \"The user's request concerns exchange via app.\", \"candidate_a11c56\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_907eea\": \"The user's request concerns exchange rate.\", \"candidate_8e3261\": \"The user's request concerns top up reverted.\", \"candidate_5ee305\": \"The user's request concerns card delivery estimate.\", \"candidate_128697\": \"The user's request concerns failed transfer.\", \"candidate_416827\": \"The user's request concerns extra charge on statement.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9612c9\": 0.0, \"candidate_0f7292\": 0.0, \"candidate_a11c56\": 1.0, \"candidate_907eea\": 0.0, \"candidate_8e3261\": 0.0, \"candidate_5ee305\": 0.0, \"candidate_128697\": 0.0, \"candidate_416827\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3f195cd07fe40349686f78a8:choice", "group_id": "fresh12:banking77:3f195cd07fe40349686f78a8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is a transaction from my account that I don't recognize can you trace back the information so I can make sure it's something I did or not?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c838be\": \"The user's request concerns verify source of funds.\", \"candidate_6b0836\": \"The user's request concerns card acceptance.\", \"candidate_fe94cf\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_199e61\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_43809d\": \"The user's request concerns getting spare card.\", \"candidate_e7f112\": \"The user's request concerns request refund.\", \"candidate_f1cc3e\": \"The user's request concerns top up failed.\", \"candidate_081b99\": \"The user's request concerns transfer timing.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c838be\": 0.0, \"candidate_6b0836\": 0.0, \"candidate_fe94cf\": 0.0, \"candidate_199e61\": 1.0, \"candidate_43809d\": 0.0, \"candidate_e7f112\": 0.0, \"candidate_f1cc3e\": 0.0, \"candidate_081b99\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8ca143b8a6211b08ce8d532d:choice", "group_id": "fresh12:banking77:8ca143b8a6211b08ce8d532d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I am concerned about the security in my account and would like to make a dispute.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fd6565\": \"The user's request concerns Refund not showing up.\", \"candidate_fe1017\": \"The user's request concerns lost or stolen phone.\", \"candidate_4c79b0\": \"The user's request concerns compromised card.\", \"candidate_411fe9\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_03aad6\": \"The user's request concerns transfer into account.\", \"candidate_fc0d08\": \"The user's request concerns declined cash withdrawal.\", \"candidate_3c6643\": \"The user's request concerns age limit.\", \"candidate_b8c286\": \"The user's request concerns declined card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fd6565\": 0.0, \"candidate_fe1017\": 0.0, \"candidate_4c79b0\": 0.0, \"candidate_411fe9\": 1.0, \"candidate_03aad6\": 0.0, \"candidate_fc0d08\": 0.0, \"candidate_3c6643\": 0.0, \"candidate_b8c286\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0abeb6e03a27daaa14546c08:choice", "group_id": "fresh12:banking77:0abeb6e03a27daaa14546c08", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is a direct debit that's not mine.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f6629c\": \"The user's request concerns getting virtual card.\", \"candidate_b122db\": \"The user's request concerns extra charge on statement.\", \"candidate_e7198e\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_b2a721\": \"The user's request concerns edit personal details.\", \"candidate_06b88a\": \"The user's request concerns verify top up.\", \"candidate_1aec5e\": \"The user's request concerns transfer timing.\", \"candidate_4477bf\": \"The user's request concerns cancel transfer.\", \"candidate_7fb247\": \"The user's request concerns get physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f6629c\": 0.0, \"candidate_b122db\": 0.0, \"candidate_e7198e\": 1.0, \"candidate_b2a721\": 0.0, \"candidate_06b88a\": 0.0, \"candidate_1aec5e\": 0.0, \"candidate_4477bf\": 0.0, \"candidate_7fb247\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1524e36bd6e88c8039b809cd:choice", "group_id": "fresh12:banking77:1524e36bd6e88c8039b809cd", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is a suspicious direct debit payment.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e829d3\": \"The user's request concerns card not working.\", \"candidate_23b998\": \"The user's request concerns terminate account.\", \"candidate_3c3cbb\": \"The user's request concerns virtual card not working.\", \"candidate_fd90f0\": \"The user's request concerns declined transfer.\", \"candidate_85f845\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_7a8fae\": \"The user's request concerns declined card payment.\", \"candidate_b73d2a\": \"The user's request concerns wrong amount of cash received.\", \"candidate_926e24\": \"The user's request concerns direct debit payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e829d3\": 0.0, \"candidate_23b998\": 0.0, \"candidate_3c3cbb\": 0.0, \"candidate_fd90f0\": 0.0, \"candidate_85f845\": 0.0, \"candidate_7a8fae\": 0.0, \"candidate_b73d2a\": 0.0, \"candidate_926e24\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e6571b3a080d0869b120008d:choice", "group_id": "fresh12:banking77:e6571b3a080d0869b120008d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I have a strange direct debit in my statement\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7d88e1\": \"The user's request concerns transaction charged twice.\", \"candidate_67b944\": \"The user's request concerns get disposable virtual card.\", \"candidate_dbd974\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_8b987c\": \"The user's request concerns atm support.\", \"candidate_96ba09\": \"The user's request concerns pending cash withdrawal.\", \"candidate_859d57\": \"The user's request concerns fiat currency support.\", \"candidate_9694d7\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_a8e18f\": \"The user's request concerns card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7d88e1\": 0.0, \"candidate_67b944\": 0.0, \"candidate_dbd974\": 1.0, \"candidate_8b987c\": 0.0, \"candidate_96ba09\": 0.0, \"candidate_859d57\": 0.0, \"candidate_9694d7\": 0.0, \"candidate_a8e18f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:62c72c4dd8c5d3be874dd5c0:choice", "group_id": "fresh12:banking77:62c72c4dd8c5d3be874dd5c0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need to reset my passcode. How do I do it?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_936989\": \"The user's request concerns transfer fee charged.\", \"candidate_0c109e\": \"The user's request concerns passcode forgotten.\", \"candidate_aeaefb\": \"The user's request concerns get physical card.\", \"candidate_dc7de3\": \"The user's request concerns age limit.\", \"candidate_32a40c\": \"The user's request concerns reverted card payment?.\", \"candidate_993d8f\": \"The user's request concerns top up failed.\", \"candidate_703c25\": \"The user's request concerns automatic top up.\", \"candidate_cf6da2\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_936989\": 0.0, \"candidate_0c109e\": 1.0, \"candidate_aeaefb\": 0.0, \"candidate_dc7de3\": 0.0, \"candidate_32a40c\": 0.0, \"candidate_993d8f\": 0.0, \"candidate_703c25\": 0.0, \"candidate_cf6da2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4de6d65604e9e2d8970744c5:choice", "group_id": "fresh12:banking77:4de6d65604e9e2d8970744c5", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can you tell me what to do to reset my passcode?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5c7a4c\": \"The user's request concerns lost or stolen phone.\", \"candidate_fc9461\": \"The user's request concerns get physical card.\", \"candidate_b009fe\": \"The user's request concerns passcode forgotten.\", \"candidate_967758\": \"The user's request concerns transfer fee charged.\", \"candidate_28bb62\": \"The user's request concerns change pin.\", \"candidate_7cc50d\": \"The user's request concerns compromised card.\", \"candidate_19e1f9\": \"The user's request concerns pending top up.\", \"candidate_68a03f\": \"The user's request concerns request refund.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5c7a4c\": 0.0, \"candidate_fc9461\": 0.0, \"candidate_b009fe\": 1.0, \"candidate_967758\": 0.0, \"candidate_28bb62\": 0.0, \"candidate_7cc50d\": 0.0, \"candidate_19e1f9\": 0.0, \"candidate_68a03f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b95d325bbeb9c78255338994:choice", "group_id": "fresh12:banking77:b95d325bbeb9c78255338994", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I retrieve my passcode?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5b3c82\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_6e6738\": \"The user's request concerns card not working.\", \"candidate_684d19\": \"The user's request concerns lost or stolen phone.\", \"candidate_c8d72e\": \"The user's request concerns receiving money.\", \"candidate_4aace9\": \"The user's request concerns get physical card.\", \"candidate_bbe6ad\": \"The user's request concerns automatic top up.\", \"candidate_bd13a8\": \"The user's request concerns change pin.\", \"candidate_43da15\": \"The user's request concerns passcode forgotten.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5b3c82\": 0.0, \"candidate_6e6738\": 0.0, \"candidate_684d19\": 0.0, \"candidate_c8d72e\": 0.0, \"candidate_4aace9\": 0.0, \"candidate_bbe6ad\": 0.0, \"candidate_bd13a8\": 0.0, \"candidate_43da15\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e3e631fa92a58e95e5b3f542:choice", "group_id": "fresh12:banking77:e3e631fa92a58e95e5b3f542", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I resent my passcode?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c09eb9\": \"The user's request concerns declined cash withdrawal.\", \"candidate_360b3a\": \"The user's request concerns declined transfer.\", \"candidate_02e1a3\": \"The user's request concerns verify my identity.\", \"candidate_d9d1c6\": \"The user's request concerns card payment fee charged.\", \"candidate_8794d5\": \"The user's request concerns card payment not recognised.\", \"candidate_7ae51d\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_ec5590\": \"The user's request concerns passcode forgotten.\", \"candidate_9cb533\": \"The user's request concerns edit personal details.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c09eb9\": 0.0, \"candidate_360b3a\": 0.0, \"candidate_02e1a3\": 0.0, \"candidate_d9d1c6\": 0.0, \"candidate_8794d5\": 0.0, \"candidate_7ae51d\": 0.0, \"candidate_ec5590\": 1.0, \"candidate_9cb533\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c9228bf90d9f1b41bbd924c1:choice", "group_id": "fresh12:banking77:c9228bf90d9f1b41bbd924c1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I have forgotten my password.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_202ab8\": \"The user's request concerns lost or stolen phone.\", \"candidate_f91fd6\": \"The user's request concerns compromised card.\", \"candidate_934478\": \"The user's request concerns top up limits.\", \"candidate_736e96\": \"The user's request concerns card linking.\", \"candidate_f89046\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_fc7985\": \"The user's request concerns top up by card charge.\", \"candidate_8e2d97\": \"The user's request concerns passcode forgotten.\", \"candidate_c16668\": \"The user's request concerns automatic top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_202ab8\": 0.0, \"candidate_f91fd6\": 0.0, \"candidate_934478\": 0.0, \"candidate_736e96\": 0.0, \"candidate_f89046\": 0.0, \"candidate_fc7985\": 0.0, \"candidate_8e2d97\": 1.0, \"candidate_c16668\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c9bc82ae123d62feecb206e3:choice", "group_id": "fresh12:banking77:c9bc82ae123d62feecb206e3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What steps do I take to receive another passcode?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9042b9\": \"The user's request concerns card not working.\", \"candidate_8d7a98\": \"The user's request concerns Refund not showing up.\", \"candidate_d6fc88\": \"The user's request concerns reverted card payment?.\", \"candidate_cd5322\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_a60309\": \"The user's request concerns declined card payment.\", \"candidate_5179d7\": \"The user's request concerns card linking.\", \"candidate_03bc0f\": \"The user's request concerns pending cash withdrawal.\", \"candidate_5a8f2b\": \"The user's request concerns passcode forgotten.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9042b9\": 0.0, \"candidate_8d7a98\": 0.0, \"candidate_d6fc88\": 0.0, \"candidate_cd5322\": 0.0, \"candidate_a60309\": 0.0, \"candidate_5179d7\": 0.0, \"candidate_03bc0f\": 0.0, \"candidate_5a8f2b\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:fc6e332ec09d077d850833d2:choice", "group_id": "fresh12:banking77:fc6e332ec09d077d850833d2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why could I not choose cash at the ATM?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2a38a0\": \"The user's request concerns compromised card.\", \"candidate_da67c9\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_cb08c5\": \"The user's request concerns declined cash withdrawal.\", \"candidate_45fe63\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_d2ac0f\": \"The user's request concerns why verify identity.\", \"candidate_0c3906\": \"The user's request concerns virtual card not working.\", \"candidate_0a43fa\": \"The user's request concerns edit personal details.\", \"candidate_23bc43\": \"The user's request concerns pending transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2a38a0\": 0.0, \"candidate_da67c9\": 0.0, \"candidate_cb08c5\": 1.0, \"candidate_45fe63\": 0.0, \"candidate_d2ac0f\": 0.0, \"candidate_0c3906\": 0.0, \"candidate_0a43fa\": 0.0, \"candidate_23bc43\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f7fda2a01cecc1d58c6c7c8b:choice", "group_id": "fresh12:banking77:f7fda2a01cecc1d58c6c7c8b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why didn't the ATM give me my money?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dba46a\": \"The user's request concerns change pin.\", \"candidate_467f51\": \"The user's request concerns automatic top up.\", \"candidate_ae5fc1\": \"The user's request concerns declined cash withdrawal.\", \"candidate_1bee1a\": \"The user's request concerns top up limits.\", \"candidate_1959db\": \"The user's request concerns why verify identity.\", \"candidate_ea75b5\": \"The user's request concerns top up failed.\", \"candidate_3cd8fd\": \"The user's request concerns age limit.\", \"candidate_787502\": \"The user's request concerns compromised card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dba46a\": 0.0, \"candidate_467f51\": 0.0, \"candidate_ae5fc1\": 1.0, \"candidate_1bee1a\": 0.0, \"candidate_1959db\": 0.0, \"candidate_ea75b5\": 0.0, \"candidate_3cd8fd\": 0.0, \"candidate_787502\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:13e0d078403beeacb9efb473:choice", "group_id": "fresh12:banking77:13e0d078403beeacb9efb473", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How come I can't take money out of the ATM?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_96acc5\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_eddcd1\": \"The user's request concerns transfer timing.\", \"candidate_e294eb\": \"The user's request concerns declined cash withdrawal.\", \"candidate_d706d0\": \"The user's request concerns order physical card.\", \"candidate_8f9bba\": \"The user's request concerns transfer fee charged.\", \"candidate_8fd224\": \"The user's request concerns contactless not working.\", \"candidate_14b739\": \"The user's request concerns get disposable virtual card.\", \"candidate_7337b0\": \"The user's request concerns supported cards and currencies.\"}}}}", "targets_json": "{\"decision\": {\"candidate_96acc5\": 0.0, \"candidate_eddcd1\": 0.0, \"candidate_e294eb\": 1.0, \"candidate_d706d0\": 0.0, \"candidate_8f9bba\": 0.0, \"candidate_8fd224\": 0.0, \"candidate_14b739\": 0.0, \"candidate_7337b0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f0da9883a7d493cdcb446ecc:choice", "group_id": "fresh12:banking77:f0da9883a7d493cdcb446ecc", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a reason that the ATM declined my transaction?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f879d2\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_53acdd\": \"The user's request concerns reverted card payment?.\", \"candidate_1c9d09\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_2599a3\": \"The user's request concerns card about to expire.\", \"candidate_fe903e\": \"The user's request concerns declined cash withdrawal.\", \"candidate_b023d7\": \"The user's request concerns compromised card.\", \"candidate_d1a757\": \"The user's request concerns supported cards and currencies.\", \"candidate_a45781\": \"The user's request concerns card delivery estimate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f879d2\": 0.0, \"candidate_53acdd\": 0.0, \"candidate_1c9d09\": 0.0, \"candidate_2599a3\": 0.0, \"candidate_fe903e\": 1.0, \"candidate_b023d7\": 0.0, \"candidate_d1a757\": 0.0, \"candidate_a45781\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e7c9dc6727be4671f8f3a750:choice", "group_id": "fresh12:banking77:e7c9dc6727be4671f8f3a750", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The ATM keeps declining my card. I tried at a couple different ATMs. Can you tell me if there's a problem with my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_62c6e6\": \"The user's request concerns order physical card.\", \"candidate_8fa987\": \"The user's request concerns top up limits.\", \"candidate_1e4f5f\": \"The user's request concerns receiving money.\", \"candidate_1d3b96\": \"The user's request concerns exchange via app.\", \"candidate_fadb2b\": \"The user's request concerns disposable card limits.\", \"candidate_bd1092\": \"The user's request concerns pending top up.\", \"candidate_6d6713\": \"The user's request concerns declined cash withdrawal.\", \"candidate_d395e2\": \"The user's request concerns top up by bank transfer charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_62c6e6\": 0.0, \"candidate_8fa987\": 0.0, \"candidate_1e4f5f\": 0.0, \"candidate_1d3b96\": 0.0, \"candidate_fadb2b\": 0.0, \"candidate_bd1092\": 0.0, \"candidate_6d6713\": 1.0, \"candidate_d395e2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9f703bb5a79c0661c7e03740:choice", "group_id": "fresh12:banking77:9f703bb5a79c0661c7e03740", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My ATM transaction was unsuccessful.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7a9f6f\": \"The user's request concerns cancel transfer.\", \"candidate_b40d3f\": \"The user's request concerns transfer not received by recipient.\", \"candidate_ad9344\": \"The user's request concerns atm support.\", \"candidate_29915c\": \"The user's request concerns topping up by card.\", \"candidate_a5e748\": \"The user's request concerns declined cash withdrawal.\", \"candidate_c888b1\": \"The user's request concerns pin blocked.\", \"candidate_55359b\": \"The user's request concerns card swallowed.\", \"candidate_458a60\": \"The user's request concerns unable to verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7a9f6f\": 0.0, \"candidate_b40d3f\": 0.0, \"candidate_ad9344\": 0.0, \"candidate_29915c\": 0.0, \"candidate_a5e748\": 1.0, \"candidate_c888b1\": 0.0, \"candidate_55359b\": 0.0, \"candidate_458a60\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:cd9aa23732b243433b60dd45:choice", "group_id": "fresh12:banking77:cd9aa23732b243433b60dd45", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do i finalize a card payment\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_28e4d7\": \"The user's request concerns pending top up.\", \"candidate_00787b\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_49a609\": \"The user's request concerns card acceptance.\", \"candidate_bfbcfb\": \"The user's request concerns country support.\", \"candidate_8b7e68\": \"The user's request concerns extra charge on statement.\", \"candidate_b913b9\": \"The user's request concerns pending card payment.\", \"candidate_dfba34\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_f3ab42\": \"The user's request concerns virtual card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_28e4d7\": 0.0, \"candidate_00787b\": 0.0, \"candidate_49a609\": 0.0, \"candidate_bfbcfb\": 0.0, \"candidate_8b7e68\": 0.0, \"candidate_b913b9\": 1.0, \"candidate_dfba34\": 0.0, \"candidate_f3ab42\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0b4476b6a1756bbde5a1774c:choice", "group_id": "fresh12:banking77:0b4476b6a1756bbde5a1774c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why is there a pending transaction on my card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_305461\": \"The user's request concerns pending transfer.\", \"candidate_218ec8\": \"The user's request concerns declined transfer.\", \"candidate_1e795e\": \"The user's request concerns pending card payment.\", \"candidate_6b40d9\": \"The user's request concerns top up by cash or cheque.\", \"candidate_13b581\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_da86a8\": \"The user's request concerns declined cash withdrawal.\", \"candidate_044d81\": \"The user's request concerns exchange charge.\", \"candidate_032753\": \"The user's request concerns automatic top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_305461\": 0.0, \"candidate_218ec8\": 0.0, \"candidate_1e795e\": 1.0, \"candidate_6b40d9\": 0.0, \"candidate_13b581\": 0.0, \"candidate_da86a8\": 0.0, \"candidate_044d81\": 0.0, \"candidate_032753\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:301b5fc76642d63867bf81f6:choice", "group_id": "fresh12:banking77:301b5fc76642d63867bf81f6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long does an item I bought show as pending?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_22bcfd\": \"The user's request concerns exchange via app.\", \"candidate_b24d8a\": \"The user's request concerns pending card payment.\", \"candidate_3727f3\": \"The user's request concerns activate my card.\", \"candidate_3ab7ac\": \"The user's request concerns pin blocked.\", \"candidate_298b6c\": \"The user's request concerns top up reverted.\", \"candidate_715d1c\": \"The user's request concerns order physical card.\", \"candidate_5a6476\": \"The user's request concerns card delivery estimate.\", \"candidate_43ae0a\": \"The user's request concerns topping up by card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_22bcfd\": 0.0, \"candidate_b24d8a\": 1.0, \"candidate_3727f3\": 0.0, \"candidate_3ab7ac\": 0.0, \"candidate_298b6c\": 0.0, \"candidate_715d1c\": 0.0, \"candidate_5a6476\": 0.0, \"candidate_43ae0a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:85171f4fd0a7f97058cf0ae9:choice", "group_id": "fresh12:banking77:85171f4fd0a7f97058cf0ae9", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card payment is showing pending.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9276da\": \"The user's request concerns declined cash withdrawal.\", \"candidate_def4a6\": \"The user's request concerns transfer fee charged.\", \"candidate_7ec0d7\": \"The user's request concerns lost or stolen card.\", \"candidate_0f756e\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_cb461e\": \"The user's request concerns declined transfer.\", \"candidate_971c59\": \"The user's request concerns pending card payment.\", \"candidate_42aa19\": \"The user's request concerns pending transfer.\", \"candidate_29b2ee\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9276da\": 0.0, \"candidate_def4a6\": 0.0, \"candidate_7ec0d7\": 0.0, \"candidate_0f756e\": 0.0, \"candidate_cb461e\": 0.0, \"candidate_971c59\": 1.0, \"candidate_42aa19\": 0.0, \"candidate_29b2ee\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d4335d8b8aa1aaa981e01ada:choice", "group_id": "fresh12:banking77:d4335d8b8aa1aaa981e01ada", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long do card transactions take to clear?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f18d7f\": \"The user's request concerns verify top up.\", \"candidate_cb558f\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_60511c\": \"The user's request concerns unable to verify identity.\", \"candidate_4c4751\": \"The user's request concerns fiat currency support.\", \"candidate_d6f1d1\": \"The user's request concerns exchange charge.\", \"candidate_6228ef\": \"The user's request concerns pending card payment.\", \"candidate_4ae8a7\": \"The user's request concerns terminate account.\", \"candidate_609206\": \"The user's request concerns top up limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f18d7f\": 0.0, \"candidate_cb558f\": 0.0, \"candidate_60511c\": 0.0, \"candidate_4c4751\": 0.0, \"candidate_d6f1d1\": 0.0, \"candidate_6228ef\": 1.0, \"candidate_4ae8a7\": 0.0, \"candidate_609206\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f9add4d4795503544b51e88e:choice", "group_id": "fresh12:banking77:f9add4d4795503544b51e88e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I made a purchase and it says it's pending. What does that mean?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ff0188\": \"The user's request concerns transfer fee charged.\", \"candidate_5d3dcd\": \"The user's request concerns lost or stolen card.\", \"candidate_a12042\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_138096\": \"The user's request concerns pending card payment.\", \"candidate_8f6a5a\": \"The user's request concerns extra charge on statement.\", \"candidate_66a03f\": \"The user's request concerns exchange via app.\", \"candidate_b51b79\": \"The user's request concerns declined transfer.\", \"candidate_447031\": \"The user's request concerns apple pay or google pay.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ff0188\": 0.0, \"candidate_5d3dcd\": 0.0, \"candidate_a12042\": 0.0, \"candidate_138096\": 1.0, \"candidate_8f6a5a\": 0.0, \"candidate_66a03f\": 0.0, \"candidate_b51b79\": 0.0, \"candidate_447031\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d14deac5e4be7d775aa2586e:choice", "group_id": "fresh12:banking77:d14deac5e4be7d775aa2586e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My phone is lost.  What do I do to prevent someone from using the app?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c37ef2\": \"The user's request concerns transaction charged twice.\", \"candidate_8d3124\": \"The user's request concerns reverted card payment?.\", \"candidate_cd8210\": \"The user's request concerns cancel transfer.\", \"candidate_826eb6\": \"The user's request concerns pending card payment.\", \"candidate_3a317c\": \"The user's request concerns declined card payment.\", \"candidate_28834a\": \"The user's request concerns topping up by card.\", \"candidate_f7447d\": \"The user's request concerns lost or stolen phone.\", \"candidate_2a74fa\": \"The user's request concerns get physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c37ef2\": 0.0, \"candidate_8d3124\": 0.0, \"candidate_cd8210\": 0.0, \"candidate_826eb6\": 0.0, \"candidate_3a317c\": 0.0, \"candidate_28834a\": 0.0, \"candidate_f7447d\": 1.0, \"candidate_2a74fa\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:180a9532f1f2b7d4f0f1ac36:choice", "group_id": "fresh12:banking77:180a9532f1f2b7d4f0f1ac36", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Someone attacked me yesterday and took my things, so I am unable to use the app.  I am in need of some help.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8d609b\": \"The user's request concerns virtual card not working.\", \"candidate_466e6e\": \"The user's request concerns request refund.\", \"candidate_98a3c1\": \"The user's request concerns Refund not showing up.\", \"candidate_2e1208\": \"The user's request concerns card about to expire.\", \"candidate_8368d7\": \"The user's request concerns lost or stolen phone.\", \"candidate_44d96b\": \"The user's request concerns extra charge on statement.\", \"candidate_a3c91b\": \"The user's request concerns card arrival.\", \"candidate_d7612b\": \"The user's request concerns top up by card charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8d609b\": 0.0, \"candidate_466e6e\": 0.0, \"candidate_98a3c1\": 0.0, \"candidate_2e1208\": 0.0, \"candidate_8368d7\": 1.0, \"candidate_44d96b\": 0.0, \"candidate_a3c91b\": 0.0, \"candidate_d7612b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9288c8cb4d966823f6e8aae7:choice", "group_id": "fresh12:banking77:9288c8cb4d966823f6e8aae7", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I can't find my phone but it has all my credit card information.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_379a71\": \"The user's request concerns pending card payment.\", \"candidate_e3755f\": \"The user's request concerns pending top up.\", \"candidate_04fc92\": \"The user's request concerns declined card payment.\", \"candidate_b8050b\": \"The user's request concerns apple pay or google pay.\", \"candidate_86191a\": \"The user's request concerns change pin.\", \"candidate_b3f02a\": \"The user's request concerns disposable card limits.\", \"candidate_91686c\": \"The user's request concerns lost or stolen phone.\", \"candidate_bb9036\": \"The user's request concerns getting spare card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_379a71\": 0.0, \"candidate_e3755f\": 0.0, \"candidate_04fc92\": 0.0, \"candidate_b8050b\": 0.0, \"candidate_86191a\": 0.0, \"candidate_b3f02a\": 0.0, \"candidate_91686c\": 1.0, \"candidate_bb9036\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:bc01434ad941d140a694fc07:choice", "group_id": "fresh12:banking77:bc01434ad941d140a694fc07", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My phone is not with me at the moment so I will not be able to access the app.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3ba62e\": \"The user's request concerns reverted card payment?.\", \"candidate_c89d31\": \"The user's request concerns virtual card not working.\", \"candidate_451f9d\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_738f35\": \"The user's request concerns card payment fee charged.\", \"candidate_87f481\": \"The user's request concerns lost or stolen phone.\", \"candidate_e863af\": \"The user's request concerns transfer into account.\", \"candidate_f69c85\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_966227\": \"The user's request concerns topping up by card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3ba62e\": 0.0, \"candidate_c89d31\": 0.0, \"candidate_451f9d\": 0.0, \"candidate_738f35\": 0.0, \"candidate_87f481\": 1.0, \"candidate_e863af\": 0.0, \"candidate_f69c85\": 0.0, \"candidate_966227\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:571b86b0c5762a15da0c2bad:choice", "group_id": "fresh12:banking77:571b86b0c5762a15da0c2bad", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I think my phone is either lost or stolen.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7473d2\": \"The user's request concerns declined transfer.\", \"candidate_e07ed4\": \"The user's request concerns lost or stolen card.\", \"candidate_09bce1\": \"The user's request concerns declined cash withdrawal.\", \"candidate_105fda\": \"The user's request concerns lost or stolen phone.\", \"candidate_935fc2\": \"The user's request concerns top up by card charge.\", \"candidate_0ae827\": \"The user's request concerns transfer fee charged.\", \"candidate_39b617\": \"The user's request concerns cash withdrawal charge.\", \"candidate_c1380b\": \"The user's request concerns verify my identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7473d2\": 0.0, \"candidate_e07ed4\": 0.0, \"candidate_09bce1\": 0.0, \"candidate_105fda\": 1.0, \"candidate_935fc2\": 0.0, \"candidate_0ae827\": 0.0, \"candidate_39b617\": 0.0, \"candidate_c1380b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2d296c218135e075f23321a6:choice", "group_id": "fresh12:banking77:2d296c218135e075f23321a6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My phone is not on me.  How can I use the app?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ab3e16\": \"The user's request concerns extra charge on statement.\", \"candidate_33ba64\": \"The user's request concerns edit personal details.\", \"candidate_257d9b\": \"The user's request concerns card payment not recognised.\", \"candidate_75eda0\": \"The user's request concerns lost or stolen phone.\", \"candidate_d63dd1\": \"The user's request concerns declined transfer.\", \"candidate_bc9951\": \"The user's request concerns activate my card.\", \"candidate_1a9e88\": \"The user's request concerns disposable card limits.\", \"candidate_b7af01\": \"The user's request concerns apple pay or google pay.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ab3e16\": 0.0, \"candidate_33ba64\": 0.0, \"candidate_257d9b\": 0.0, \"candidate_75eda0\": 1.0, \"candidate_d63dd1\": 0.0, \"candidate_bc9951\": 0.0, \"candidate_1a9e88\": 0.0, \"candidate_b7af01\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7e008c27e873833ffaaf98af:choice", "group_id": "fresh12:banking77:7e008c27e873833ffaaf98af", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I want to reverse a purchase. Can I cancel it?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c6a0de\": \"The user's request concerns apple pay or google pay.\", \"candidate_6be644\": \"The user's request concerns declined cash withdrawal.\", \"candidate_5e83de\": \"The user's request concerns request refund.\", \"candidate_fdb3c4\": \"The user's request concerns pending cash withdrawal.\", \"candidate_c9cb67\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_a93c99\": \"The user's request concerns top up reverted.\", \"candidate_f2a5f6\": \"The user's request concerns card not working.\", \"candidate_7a79cf\": \"The user's request concerns compromised card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c6a0de\": 0.0, \"candidate_6be644\": 0.0, \"candidate_5e83de\": 1.0, \"candidate_fdb3c4\": 0.0, \"candidate_c9cb67\": 0.0, \"candidate_a93c99\": 0.0, \"candidate_f2a5f6\": 0.0, \"candidate_7a79cf\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ce602a868f4f539e900689f8:choice", "group_id": "fresh12:banking77:ce602a868f4f539e900689f8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I want this amount reversed out of my account.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a50bac\": \"The user's request concerns get physical card.\", \"candidate_ce4f56\": \"The user's request concerns request refund.\", \"candidate_1dbc4a\": \"The user's request concerns why verify identity.\", \"candidate_d1c77c\": \"The user's request concerns age limit.\", \"candidate_5945fb\": \"The user's request concerns order physical card.\", \"candidate_f79d49\": \"The user's request concerns card acceptance.\", \"candidate_ca31ee\": \"The user's request concerns passcode forgotten.\", \"candidate_d094f2\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a50bac\": 0.0, \"candidate_ce4f56\": 1.0, \"candidate_1dbc4a\": 0.0, \"candidate_d1c77c\": 0.0, \"candidate_5945fb\": 0.0, \"candidate_f79d49\": 0.0, \"candidate_ca31ee\": 0.0, \"candidate_d094f2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8e5b67f7e73063c9e0868b44:choice", "group_id": "fresh12:banking77:8e5b67f7e73063c9e0868b44", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I have a refund?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_92bf6f\": \"The user's request concerns card delivery estimate.\", \"candidate_3082d4\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_66a320\": \"The user's request concerns country support.\", \"candidate_85111d\": \"The user's request concerns card acceptance.\", \"candidate_2d3b16\": \"The user's request concerns request refund.\", \"candidate_a3f8a9\": \"The user's request concerns pending top up.\", \"candidate_16558a\": \"The user's request concerns exchange charge.\", \"candidate_63137c\": \"The user's request concerns activate my card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_92bf6f\": 0.0, \"candidate_3082d4\": 0.0, \"candidate_66a320\": 0.0, \"candidate_85111d\": 0.0, \"candidate_2d3b16\": 1.0, \"candidate_a3f8a9\": 0.0, \"candidate_16558a\": 0.0, \"candidate_63137c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ef1ec2878b238a653c370ba0:choice", "group_id": "fresh12:banking77:ef1ec2878b238a653c370ba0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I bought this item and was charged the wrong amount can I get a refund?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ab1f3a\": \"The user's request concerns terminate account.\", \"candidate_3933f3\": \"The user's request concerns verify my identity.\", \"candidate_090b71\": \"The user's request concerns atm support.\", \"candidate_3d38c5\": \"The user's request concerns transaction charged twice.\", \"candidate_810ac6\": \"The user's request concerns extra charge on statement.\", \"candidate_7ada9d\": \"The user's request concerns fiat currency support.\", \"candidate_6c1eae\": \"The user's request concerns request refund.\", \"candidate_703295\": \"The user's request concerns card arrival.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ab1f3a\": 0.0, \"candidate_3933f3\": 0.0, \"candidate_090b71\": 0.0, \"candidate_3d38c5\": 0.0, \"candidate_810ac6\": 0.0, \"candidate_7ada9d\": 0.0, \"candidate_6c1eae\": 1.0, \"candidate_703295\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9464f7fb1ec4e5037e665d6f:choice", "group_id": "fresh12:banking77:9464f7fb1ec4e5037e665d6f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I would appreciate it if I could get an item refunded\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9b40ac\": \"The user's request concerns edit personal details.\", \"candidate_04646a\": \"The user's request concerns request refund.\", \"candidate_3e9ca1\": \"The user's request concerns reverted card payment?.\", \"candidate_978213\": \"The user's request concerns atm support.\", \"candidate_2d8ae3\": \"The user's request concerns terminate account.\", \"candidate_d87394\": \"The user's request concerns automatic top up.\", \"candidate_c26a5e\": \"The user's request concerns card about to expire.\", \"candidate_a29c17\": \"The user's request concerns order physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9b40ac\": 0.0, \"candidate_04646a\": 1.0, \"candidate_3e9ca1\": 0.0, \"candidate_978213\": 0.0, \"candidate_2d8ae3\": 0.0, \"candidate_d87394\": 0.0, \"candidate_c26a5e\": 0.0, \"candidate_a29c17\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3ddd594ea3ed16bdf1869757:choice", "group_id": "fresh12:banking77:3ddd594ea3ed16bdf1869757", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I am unhappy with my purchase, how do I cancel the order?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_072204\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_f050b5\": \"The user's request concerns country support.\", \"candidate_aafa1f\": \"The user's request concerns disposable card limits.\", \"candidate_659dcb\": \"The user's request concerns request refund.\", \"candidate_312598\": \"The user's request concerns visa or mastercard.\", \"candidate_7e7e67\": \"The user's request concerns atm support.\", \"candidate_097cb0\": \"The user's request concerns automatic top up.\", \"candidate_cd48c6\": \"The user's request concerns wrong exchange rate for cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_072204\": 0.0, \"candidate_f050b5\": 0.0, \"candidate_aafa1f\": 0.0, \"candidate_659dcb\": 1.0, \"candidate_312598\": 0.0, \"candidate_7e7e67\": 0.0, \"candidate_097cb0\": 0.0, \"candidate_cd48c6\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:037161a597135f559488ef5b:choice", "group_id": "fresh12:banking77:037161a597135f559488ef5b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"why was my transfer declined\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3f5f87\": \"The user's request concerns disposable card limits.\", \"candidate_0c4081\": \"The user's request concerns passcode forgotten.\", \"candidate_ec9082\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_4b5312\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_37e82b\": \"The user's request concerns top up by card charge.\", \"candidate_96be1f\": \"The user's request concerns declined transfer.\", \"candidate_702ec4\": \"The user's request concerns receiving money.\", \"candidate_50a399\": \"The user's request concerns direct debit payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3f5f87\": 0.0, \"candidate_0c4081\": 0.0, \"candidate_ec9082\": 0.0, \"candidate_4b5312\": 0.0, \"candidate_37e82b\": 0.0, \"candidate_96be1f\": 1.0, \"candidate_702ec4\": 0.0, \"candidate_50a399\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:803ae923df55036ad6d3ec88:choice", "group_id": "fresh12:banking77:803ae923df55036ad6d3ec88", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why did my transfer decline\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3ea199\": \"The user's request concerns transaction charged twice.\", \"candidate_df74f3\": \"The user's request concerns declined transfer.\", \"candidate_0dba24\": \"The user's request concerns transfer timing.\", \"candidate_a74803\": \"The user's request concerns top up by cash or cheque.\", \"candidate_5f2854\": \"The user's request concerns apple pay or google pay.\", \"candidate_ba69fa\": \"The user's request concerns card not working.\", \"candidate_20f625\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_f5b520\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3ea199\": 0.0, \"candidate_df74f3\": 1.0, \"candidate_0dba24\": 0.0, \"candidate_a74803\": 0.0, \"candidate_5f2854\": 0.0, \"candidate_ba69fa\": 0.0, \"candidate_20f625\": 0.0, \"candidate_f5b520\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e4320ac0f00a1b7bdd169da6:choice", "group_id": "fresh12:banking77:e4320ac0f00a1b7bdd169da6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why would my transfer be declined? I've checked that I've put in all the right details, but it is still declined.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_85824c\": \"The user's request concerns card not working.\", \"candidate_6ae660\": \"The user's request concerns wrong amount of cash received.\", \"candidate_670b9d\": \"The user's request concerns verify source of funds.\", \"candidate_e1d451\": \"The user's request concerns lost or stolen card.\", \"candidate_326913\": \"The user's request concerns declined transfer.\", \"candidate_64dba9\": \"The user's request concerns getting spare card.\", \"candidate_0a0a9b\": \"The user's request concerns transfer fee charged.\", \"candidate_2ed846\": \"The user's request concerns lost or stolen phone.\"}}}}", "targets_json": "{\"decision\": {\"candidate_85824c\": 0.0, \"candidate_6ae660\": 0.0, \"candidate_670b9d\": 0.0, \"candidate_e1d451\": 0.0, \"candidate_326913\": 1.0, \"candidate_64dba9\": 0.0, \"candidate_0a0a9b\": 0.0, \"candidate_2ed846\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:75da259f0c15bcca1809490b:choice", "group_id": "fresh12:banking77:75da259f0c15bcca1809490b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My transfer got declined.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c74ae9\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_520112\": \"The user's request concerns declined transfer.\", \"candidate_163b8e\": \"The user's request concerns cancel transfer.\", \"candidate_74108b\": \"The user's request concerns card delivery estimate.\", \"candidate_2492c8\": \"The user's request concerns request refund.\", \"candidate_df6b36\": \"The user's request concerns top up reverted.\", \"candidate_66a4af\": \"The user's request concerns declined cash withdrawal.\", \"candidate_f30984\": \"The user's request concerns verify source of funds.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c74ae9\": 0.0, \"candidate_520112\": 1.0, \"candidate_163b8e\": 0.0, \"candidate_74108b\": 0.0, \"candidate_2492c8\": 0.0, \"candidate_df6b36\": 0.0, \"candidate_66a4af\": 0.0, \"candidate_f30984\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:740fedbee48037ed99a842b6:choice", "group_id": "fresh12:banking77:740fedbee48037ed99a842b6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I am getting continuous failure for all my transfers. I have cross checked each recipient detail again and again and everything is correct.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_95a493\": \"The user's request concerns card payment not recognised.\", \"candidate_76023a\": \"The user's request concerns supported cards and currencies.\", \"candidate_25d69e\": \"The user's request concerns declined transfer.\", \"candidate_57c248\": \"The user's request concerns country support.\", \"candidate_a27c1a\": \"The user's request concerns declined card payment.\", \"candidate_466aa9\": \"The user's request concerns unable to verify identity.\", \"candidate_ae094c\": \"The user's request concerns card not working.\", \"candidate_5ee509\": \"The user's request concerns get disposable virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_95a493\": 0.0, \"candidate_76023a\": 0.0, \"candidate_25d69e\": 1.0, \"candidate_57c248\": 0.0, \"candidate_a27c1a\": 0.0, \"candidate_466aa9\": 0.0, \"candidate_ae094c\": 0.0, \"candidate_5ee509\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1f28ff813684ec368ae05970:choice", "group_id": "fresh12:banking77:1f28ff813684ec368ae05970", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My transfers are being declined. IT has always worked for me but now it's declining my card. I've tried several times.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_099c7e\": \"The user's request concerns wrong amount of cash received.\", \"candidate_f85f05\": \"The user's request concerns getting spare card.\", \"candidate_82d3ce\": \"The user's request concerns declined transfer.\", \"candidate_156a96\": \"The user's request concerns topping up by card.\", \"candidate_6b2230\": \"The user's request concerns compromised card.\", \"candidate_e3d00e\": \"The user's request concerns pending card payment.\", \"candidate_787752\": \"The user's request concerns apple pay or google pay.\", \"candidate_d3d82f\": \"The user's request concerns balance not updated after cheque or cash deposit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_099c7e\": 0.0, \"candidate_f85f05\": 0.0, \"candidate_82d3ce\": 1.0, \"candidate_156a96\": 0.0, \"candidate_6b2230\": 0.0, \"candidate_e3d00e\": 0.0, \"candidate_787752\": 0.0, \"candidate_d3d82f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d27b7685142e52e2034eaa41:choice", "group_id": "fresh12:banking77:d27b7685142e52e2034eaa41", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My refund is missing\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_103e78\": \"The user's request concerns Refund not showing up.\", \"candidate_7d12f6\": \"The user's request concerns why verify identity.\", \"candidate_239e27\": \"The user's request concerns virtual card not working.\", \"candidate_4db061\": \"The user's request concerns transaction charged twice.\", \"candidate_a1c41f\": \"The user's request concerns pending transfer.\", \"candidate_df745b\": \"The user's request concerns declined transfer.\", \"candidate_7f4fea\": \"The user's request concerns exchange rate.\", \"candidate_3ed6ad\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_103e78\": 1.0, \"candidate_7d12f6\": 0.0, \"candidate_239e27\": 0.0, \"candidate_4db061\": 0.0, \"candidate_a1c41f\": 0.0, \"candidate_df745b\": 0.0, \"candidate_7f4fea\": 0.0, \"candidate_3ed6ad\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:227e9697816bdee7f9092d77:choice", "group_id": "fresh12:banking77:227e9697816bdee7f9092d77", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I should have received a refund however it is not showing on my statement\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3c3a6b\": \"The user's request concerns Refund not showing up.\", \"candidate_6d5504\": \"The user's request concerns cash withdrawal charge.\", \"candidate_c7a009\": \"The user's request concerns automatic top up.\", \"candidate_4adcc7\": \"The user's request concerns fiat currency support.\", \"candidate_0b56a0\": \"The user's request concerns card arrival.\", \"candidate_1fdc26\": \"The user's request concerns pending card payment.\", \"candidate_348703\": \"The user's request concerns extra charge on statement.\", \"candidate_845e5f\": \"The user's request concerns why verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3c3a6b\": 1.0, \"candidate_6d5504\": 0.0, \"candidate_c7a009\": 0.0, \"candidate_4adcc7\": 0.0, \"candidate_0b56a0\": 0.0, \"candidate_1fdc26\": 0.0, \"candidate_348703\": 0.0, \"candidate_845e5f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f18c131b09b58807cc9c37db:choice", "group_id": "fresh12:banking77:f18c131b09b58807cc9c37db", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why isn't a recent refund on my statement?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6883e9\": \"The user's request concerns transfer not received by recipient.\", \"candidate_a11cb4\": \"The user's request concerns exchange charge.\", \"candidate_8ac0f5\": \"The user's request concerns card payment not recognised.\", \"candidate_b6de1b\": \"The user's request concerns getting spare card.\", \"candidate_531549\": \"The user's request concerns card swallowed.\", \"candidate_dd38be\": \"The user's request concerns Refund not showing up.\", \"candidate_404cc6\": \"The user's request concerns card delivery estimate.\", \"candidate_2aa191\": \"The user's request concerns direct debit payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6883e9\": 0.0, \"candidate_a11cb4\": 0.0, \"candidate_8ac0f5\": 0.0, \"candidate_b6de1b\": 0.0, \"candidate_531549\": 0.0, \"candidate_dd38be\": 1.0, \"candidate_404cc6\": 0.0, \"candidate_2aa191\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:03adc894fab10b94705b9474:choice", "group_id": "fresh12:banking77:03adc894fab10b94705b9474", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I still haven't seen a refund on my account\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b86aaf\": \"The user's request concerns verify my identity.\", \"candidate_6da4e9\": \"The user's request concerns apple pay or google pay.\", \"candidate_7ccee3\": \"The user's request concerns top up by card charge.\", \"candidate_006175\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_f87cae\": \"The user's request concerns getting spare card.\", \"candidate_9fdaa6\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_9d1856\": \"The user's request concerns Refund not showing up.\", \"candidate_4fb1fc\": \"The user's request concerns unable to verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b86aaf\": 0.0, \"candidate_6da4e9\": 0.0, \"candidate_7ccee3\": 0.0, \"candidate_006175\": 0.0, \"candidate_f87cae\": 0.0, \"candidate_9fdaa6\": 0.0, \"candidate_9d1856\": 1.0, \"candidate_4fb1fc\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8190cad3ad0cdb67e7a9c1dd:choice", "group_id": "fresh12:banking77:8190cad3ad0cdb67e7a9c1dd", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I asked for a refund last week but nothing has happened yet, can you help me please?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d1f9a3\": \"The user's request concerns automatic top up.\", \"candidate_4ff1ee\": \"The user's request concerns top up reverted.\", \"candidate_ce6784\": \"The user's request concerns order physical card.\", \"candidate_fd604c\": \"The user's request concerns supported cards and currencies.\", \"candidate_bd6ac0\": \"The user's request concerns reverted card payment?.\", \"candidate_b0c661\": \"The user's request concerns exchange via app.\", \"candidate_c3d2ee\": \"The user's request concerns Refund not showing up.\", \"candidate_2a0022\": \"The user's request concerns request refund.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d1f9a3\": 0.0, \"candidate_4ff1ee\": 0.0, \"candidate_ce6784\": 0.0, \"candidate_fd604c\": 0.0, \"candidate_bd6ac0\": 0.0, \"candidate_b0c661\": 0.0, \"candidate_c3d2ee\": 1.0, \"candidate_2a0022\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:722e3f668f60c46569ae7853:choice", "group_id": "fresh12:banking77:722e3f668f60c46569ae7853", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I am missing a refund.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6a9d40\": \"The user's request concerns Refund not showing up.\", \"candidate_2f125c\": \"The user's request concerns card payment not recognised.\", \"candidate_c8983d\": \"The user's request concerns pending top up.\", \"candidate_897344\": \"The user's request concerns why verify identity.\", \"candidate_8cb836\": \"The user's request concerns pin blocked.\", \"candidate_73e555\": \"The user's request concerns transfer not received by recipient.\", \"candidate_afbd63\": \"The user's request concerns declined cash withdrawal.\", \"candidate_3fa763\": \"The user's request concerns pending cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6a9d40\": 1.0, \"candidate_2f125c\": 0.0, \"candidate_c8983d\": 0.0, \"candidate_897344\": 0.0, \"candidate_8cb836\": 0.0, \"candidate_73e555\": 0.0, \"candidate_afbd63\": 0.0, \"candidate_3fa763\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8ff20ce0b0e07a70cde59d80:choice", "group_id": "fresh12:banking77:8ff20ce0b0e07a70cde59d80", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can you look at my account ans see why my card was declined\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1c014d\": \"The user's request concerns top up by cash or cheque.\", \"candidate_56e4a6\": \"The user's request concerns order physical card.\", \"candidate_eb6330\": \"The user's request concerns verify source of funds.\", \"candidate_8c68e8\": \"The user's request concerns wrong amount of cash received.\", \"candidate_287281\": \"The user's request concerns declined card payment.\", \"candidate_21c376\": \"The user's request concerns receiving money.\", \"candidate_72efff\": \"The user's request concerns transaction charged twice.\", \"candidate_814256\": \"The user's request concerns transfer fee charged.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1c014d\": 0.0, \"candidate_56e4a6\": 0.0, \"candidate_eb6330\": 0.0, \"candidate_8c68e8\": 0.0, \"candidate_287281\": 1.0, \"candidate_21c376\": 0.0, \"candidate_72efff\": 0.0, \"candidate_814256\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2a65a788ceead8a439c081ae:choice", "group_id": "fresh12:banking77:2a65a788ceead8a439c081ae", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card was rejected at the store for some reason, can you look into this for me?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e6d7c4\": \"The user's request concerns pending cash withdrawal.\", \"candidate_839876\": \"The user's request concerns topping up by card.\", \"candidate_bb6958\": \"The user's request concerns lost or stolen card.\", \"candidate_67e5a5\": \"The user's request concerns passcode forgotten.\", \"candidate_46d615\": \"The user's request concerns top up failed.\", \"candidate_8f911e\": \"The user's request concerns failed transfer.\", \"candidate_405def\": \"The user's request concerns declined card payment.\", \"candidate_dc44cc\": \"The user's request concerns exchange via app.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e6d7c4\": 0.0, \"candidate_839876\": 0.0, \"candidate_bb6958\": 0.0, \"candidate_67e5a5\": 0.0, \"candidate_46d615\": 0.0, \"candidate_8f911e\": 0.0, \"candidate_405def\": 1.0, \"candidate_dc44cc\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d3b6afe822f4ddb0b2dc78f3:choice", "group_id": "fresh12:banking77:d3b6afe822f4ddb0b2dc78f3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why was my card payment declined?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8d5eaa\": \"The user's request concerns transfer not received by recipient.\", \"candidate_257fad\": \"The user's request concerns card not working.\", \"candidate_0116aa\": \"The user's request concerns verify source of funds.\", \"candidate_0958ad\": \"The user's request concerns card delivery estimate.\", \"candidate_71cf51\": \"The user's request concerns extra charge on statement.\", \"candidate_d1ed6f\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_02d72c\": \"The user's request concerns card payment not recognised.\", \"candidate_571f9f\": \"The user's request concerns declined card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8d5eaa\": 0.0, \"candidate_257fad\": 0.0, \"candidate_0116aa\": 0.0, \"candidate_0958ad\": 0.0, \"candidate_71cf51\": 0.0, \"candidate_d1ed6f\": 0.0, \"candidate_02d72c\": 0.0, \"candidate_571f9f\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9eb0a77200b86d36bf188232:choice", "group_id": "fresh12:banking77:9eb0a77200b86d36bf188232", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why doesn't the card payment work\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_016a17\": \"The user's request concerns declined card payment.\", \"candidate_b2f45c\": \"The user's request concerns unable to verify identity.\", \"candidate_6769bc\": \"The user's request concerns pending card payment.\", \"candidate_f3a385\": \"The user's request concerns transfer timing.\", \"candidate_7fd2f7\": \"The user's request concerns card about to expire.\", \"candidate_556d9d\": \"The user's request concerns supported cards and currencies.\", \"candidate_0e970f\": \"The user's request concerns exchange rate.\", \"candidate_8da51d\": \"The user's request concerns order physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_016a17\": 1.0, \"candidate_b2f45c\": 0.0, \"candidate_6769bc\": 0.0, \"candidate_f3a385\": 0.0, \"candidate_7fd2f7\": 0.0, \"candidate_556d9d\": 0.0, \"candidate_0e970f\": 0.0, \"candidate_8da51d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0cbe9f0c47affe296bb567d5:choice", "group_id": "fresh12:banking77:0cbe9f0c47affe296bb567d5", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card was not accepted.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b2bcb1\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_4ecc63\": \"The user's request concerns declined card payment.\", \"candidate_28f5b3\": \"The user's request concerns top up by card charge.\", \"candidate_1d04dd\": \"The user's request concerns verify top up.\", \"candidate_61136b\": \"The user's request concerns pending card payment.\", \"candidate_4ad7e0\": \"The user's request concerns supported cards and currencies.\", \"candidate_85084e\": \"The user's request concerns get disposable virtual card.\", \"candidate_96e81c\": \"The user's request concerns top up failed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b2bcb1\": 0.0, \"candidate_4ecc63\": 1.0, \"candidate_28f5b3\": 0.0, \"candidate_1d04dd\": 0.0, \"candidate_61136b\": 0.0, \"candidate_4ad7e0\": 0.0, \"candidate_85084e\": 0.0, \"candidate_96e81c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1e42c2434003e7d35e8cfab4:choice", "group_id": "fresh12:banking77:1e42c2434003e7d35e8cfab4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do you fix  my card payment?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0ff226\": \"The user's request concerns exchange rate.\", \"candidate_cbb170\": \"The user's request concerns why verify identity.\", \"candidate_f8ba56\": \"The user's request concerns card about to expire.\", \"candidate_3c5a0d\": \"The user's request concerns activate my card.\", \"candidate_d4b6e5\": \"The user's request concerns transfer timing.\", \"candidate_7517d2\": \"The user's request concerns disposable card limits.\", \"candidate_08acaf\": \"The user's request concerns declined card payment.\", \"candidate_0ce7b7\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0ff226\": 0.0, \"candidate_cbb170\": 0.0, \"candidate_f8ba56\": 0.0, \"candidate_3c5a0d\": 0.0, \"candidate_d4b6e5\": 0.0, \"candidate_7517d2\": 0.0, \"candidate_08acaf\": 1.0, \"candidate_0ce7b7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:53efac201e04e8398a2ff24b:choice", "group_id": "fresh12:banking77:53efac201e04e8398a2ff24b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can I speed up a transfer?  Mine is pending.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2a076a\": \"The user's request concerns transaction charged twice.\", \"candidate_ed375d\": \"The user's request concerns pending transfer.\", \"candidate_99b1f2\": \"The user's request concerns top up failed.\", \"candidate_d581db\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_9db6b3\": \"The user's request concerns beneficiary not allowed.\", \"candidate_afb4c8\": \"The user's request concerns country support.\", \"candidate_2f94f4\": \"The user's request concerns wrong amount of cash received.\", \"candidate_d64fd2\": \"The user's request concerns top up by card charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2a076a\": 0.0, \"candidate_ed375d\": 1.0, \"candidate_99b1f2\": 0.0, \"candidate_d581db\": 0.0, \"candidate_9db6b3\": 0.0, \"candidate_afb4c8\": 0.0, \"candidate_2f94f4\": 0.0, \"candidate_d64fd2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a486e3633f7dd8dc5a2bfdd0:choice", "group_id": "fresh12:banking77:a486e3633f7dd8dc5a2bfdd0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long do transfer pend for?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0fbccb\": \"The user's request concerns pending transfer.\", \"candidate_f06868\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_f1dc4a\": \"The user's request concerns lost or stolen phone.\", \"candidate_d84185\": \"The user's request concerns top up failed.\", \"candidate_c65e71\": \"The user's request concerns pending top up.\", \"candidate_aeebf3\": \"The user's request concerns card payment fee charged.\", \"candidate_9b3c4d\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_e99f88\": \"The user's request concerns Refund not showing up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0fbccb\": 1.0, \"candidate_f06868\": 0.0, \"candidate_f1dc4a\": 0.0, \"candidate_d84185\": 0.0, \"candidate_c65e71\": 0.0, \"candidate_aeebf3\": 0.0, \"candidate_9b3c4d\": 0.0, \"candidate_e99f88\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:64a7385e26c7b63d5cd1c6ba:choice", "group_id": "fresh12:banking77:64a7385e26c7b63d5cd1c6ba", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My transfer is still coming up as pending.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4eea23\": \"The user's request concerns beneficiary not allowed.\", \"candidate_342541\": \"The user's request concerns top up failed.\", \"candidate_96f588\": \"The user's request concerns pending transfer.\", \"candidate_e35856\": \"The user's request concerns verify source of funds.\", \"candidate_9306ce\": \"The user's request concerns declined transfer.\", \"candidate_174894\": \"The user's request concerns automatic top up.\", \"candidate_0d5f0e\": \"The user's request concerns exchange via app.\", \"candidate_96b2f7\": \"The user's request concerns card swallowed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4eea23\": 0.0, \"candidate_342541\": 0.0, \"candidate_96f588\": 1.0, \"candidate_e35856\": 0.0, \"candidate_9306ce\": 0.0, \"candidate_174894\": 0.0, \"candidate_0d5f0e\": 0.0, \"candidate_96b2f7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:080d3669b3de276056071adb:choice", "group_id": "fresh12:banking77:080d3669b3de276056071adb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"When will the transfer go through?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fbed2d\": \"The user's request concerns topping up by card.\", \"candidate_9b9bd0\": \"The user's request concerns top up reverted.\", \"candidate_8a3fab\": \"The user's request concerns pending transfer.\", \"candidate_328352\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_27327e\": \"The user's request concerns apple pay or google pay.\", \"candidate_79c4fa\": \"The user's request concerns card acceptance.\", \"candidate_f48c76\": \"The user's request concerns verify top up.\", \"candidate_0fddc1\": \"The user's request concerns wrong exchange rate for cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fbed2d\": 0.0, \"candidate_9b9bd0\": 0.0, \"candidate_8a3fab\": 1.0, \"candidate_328352\": 0.0, \"candidate_27327e\": 0.0, \"candidate_79c4fa\": 0.0, \"candidate_f48c76\": 0.0, \"candidate_0fddc1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:756908d36703b73e609f6af7:choice", "group_id": "fresh12:banking77:756908d36703b73e609f6af7", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What is the cause that the transfer shows as pending?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3cd804\": \"The user's request concerns cancel transfer.\", \"candidate_740954\": \"The user's request concerns virtual card not working.\", \"candidate_74ca8d\": \"The user's request concerns card about to expire.\", \"candidate_91077a\": \"The user's request concerns pending transfer.\", \"candidate_4aa1ec\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_a36e15\": \"The user's request concerns card arrival.\", \"candidate_3b7154\": \"The user's request concerns beneficiary not allowed.\", \"candidate_ccd7e3\": \"The user's request concerns pending top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3cd804\": 0.0, \"candidate_740954\": 0.0, \"candidate_74ca8d\": 0.0, \"candidate_91077a\": 1.0, \"candidate_4aa1ec\": 0.0, \"candidate_a36e15\": 0.0, \"candidate_3b7154\": 0.0, \"candidate_ccd7e3\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:02374ecf90dbca6e32ad06c8:choice", "group_id": "fresh12:banking77:02374ecf90dbca6e32ad06c8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Someone transferred money to me and it doesn't show\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7dbf91\": \"The user's request concerns topping up by card.\", \"candidate_d0fb6e\": \"The user's request concerns order physical card.\", \"candidate_5c467c\": \"The user's request concerns card arrival.\", \"candidate_2ccc38\": \"The user's request concerns exchange charge.\", \"candidate_26d0ac\": \"The user's request concerns top up limits.\", \"candidate_b5363b\": \"The user's request concerns declined transfer.\", \"candidate_ced60b\": \"The user's request concerns pending transfer.\", \"candidate_75a501\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7dbf91\": 0.0, \"candidate_d0fb6e\": 0.0, \"candidate_5c467c\": 0.0, \"candidate_2ccc38\": 0.0, \"candidate_26d0ac\": 0.0, \"candidate_b5363b\": 0.0, \"candidate_ced60b\": 1.0, \"candidate_75a501\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:203f837cb691e452a263182e:choice", "group_id": "fresh12:banking77:203f837cb691e452a263182e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I would like to close my account.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4c3456\": \"The user's request concerns virtual card not working.\", \"candidate_f6807a\": \"The user's request concerns wrong amount of cash received.\", \"candidate_c4b012\": \"The user's request concerns cancel transfer.\", \"candidate_4fc59d\": \"The user's request concerns pending card payment.\", \"candidate_008291\": \"The user's request concerns verify top up.\", \"candidate_450a5b\": \"The user's request concerns card not working.\", \"candidate_b8a123\": \"The user's request concerns terminate account.\", \"candidate_508626\": \"The user's request concerns transaction charged twice.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4c3456\": 0.0, \"candidate_f6807a\": 0.0, \"candidate_c4b012\": 0.0, \"candidate_4fc59d\": 0.0, \"candidate_008291\": 0.0, \"candidate_450a5b\": 0.0, \"candidate_b8a123\": 1.0, \"candidate_508626\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a06f66ae5b06f57ca521bdc2:choice", "group_id": "fresh12:banking77:a06f66ae5b06f57ca521bdc2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I would like to delete my account please.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_403bd9\": \"The user's request concerns beneficiary not allowed.\", \"candidate_ffe4eb\": \"The user's request concerns get physical card.\", \"candidate_77935c\": \"The user's request concerns exchange charge.\", \"candidate_709f0e\": \"The user's request concerns country support.\", \"candidate_4c6edf\": \"The user's request concerns Refund not showing up.\", \"candidate_2b6458\": \"The user's request concerns pin blocked.\", \"candidate_f2c725\": \"The user's request concerns terminate account.\", \"candidate_e2c557\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_403bd9\": 0.0, \"candidate_ffe4eb\": 0.0, \"candidate_77935c\": 0.0, \"candidate_709f0e\": 0.0, \"candidate_4c6edf\": 0.0, \"candidate_2b6458\": 0.0, \"candidate_f2c725\": 1.0, \"candidate_e2c557\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2e4ddc80fc027558db5eb85d:choice", "group_id": "fresh12:banking77:2e4ddc80fc027558db5eb85d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Please close my account. I am unsatisfied with your service.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2bd683\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_c15e64\": \"The user's request concerns transaction charged twice.\", \"candidate_d67c7a\": \"The user's request concerns terminate account.\", \"candidate_7f3b45\": \"The user's request concerns pin blocked.\", \"candidate_eb4fa1\": \"The user's request concerns change pin.\", \"candidate_fe18c1\": \"The user's request concerns get physical card.\", \"candidate_fd88dc\": \"The user's request concerns supported cards and currencies.\", \"candidate_a5fa62\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2bd683\": 0.0, \"candidate_c15e64\": 0.0, \"candidate_d67c7a\": 1.0, \"candidate_7f3b45\": 0.0, \"candidate_eb4fa1\": 0.0, \"candidate_fe18c1\": 0.0, \"candidate_fd88dc\": 0.0, \"candidate_a5fa62\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:14f32ec753dee07267f368cc:choice", "group_id": "fresh12:banking77:14f32ec753dee07267f368cc", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"unhappy with you, how to delete account\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fc19b4\": \"The user's request concerns passcode forgotten.\", \"candidate_ae43f1\": \"The user's request concerns transfer fee charged.\", \"candidate_54dcaa\": \"The user's request concerns edit personal details.\", \"candidate_deed1d\": \"The user's request concerns getting virtual card.\", \"candidate_450d03\": \"The user's request concerns reverted card payment?.\", \"candidate_d10c08\": \"The user's request concerns cash withdrawal charge.\", \"candidate_0488f0\": \"The user's request concerns terminate account.\", \"candidate_ee7dc4\": \"The user's request concerns transfer not received by recipient.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fc19b4\": 0.0, \"candidate_ae43f1\": 0.0, \"candidate_54dcaa\": 0.0, \"candidate_deed1d\": 0.0, \"candidate_450d03\": 0.0, \"candidate_d10c08\": 0.0, \"candidate_0488f0\": 1.0, \"candidate_ee7dc4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ed83f9847c86dc80d5fd8490:choice", "group_id": "fresh12:banking77:ed83f9847c86dc80d5fd8490", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Please delete my account. I am not happy with the service from your company.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_31eebe\": \"The user's request concerns transfer into account.\", \"candidate_60f047\": \"The user's request concerns pending cash withdrawal.\", \"candidate_78a3f9\": \"The user's request concerns top up failed.\", \"candidate_a0e44d\": \"The user's request concerns terminate account.\", \"candidate_effb7e\": \"The user's request concerns wrong amount of cash received.\", \"candidate_1727f4\": \"The user's request concerns apple pay or google pay.\", \"candidate_2bb27c\": \"The user's request concerns receiving money.\", \"candidate_47661b\": \"The user's request concerns top up by cash or cheque.\"}}}}", "targets_json": "{\"decision\": {\"candidate_31eebe\": 0.0, \"candidate_60f047\": 0.0, \"candidate_78a3f9\": 0.0, \"candidate_a0e44d\": 1.0, \"candidate_effb7e\": 0.0, \"candidate_1727f4\": 0.0, \"candidate_2bb27c\": 0.0, \"candidate_47661b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:532fca89e86b889ad4f0e3dc:choice", "group_id": "fresh12:banking77:532fca89e86b889ad4f0e3dc", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I am extremely unhappy with this app and want to get rid of my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f0e04b\": \"The user's request concerns lost or stolen phone.\", \"candidate_e1dca2\": \"The user's request concerns top up limits.\", \"candidate_009ade\": \"The user's request concerns activate my card.\", \"candidate_b9213f\": \"The user's request concerns terminate account.\", \"candidate_822cec\": \"The user's request concerns passcode forgotten.\", \"candidate_c2dd39\": \"The user's request concerns top up by card charge.\", \"candidate_e4165c\": \"The user's request concerns reverted card payment?.\", \"candidate_fef432\": \"The user's request concerns get disposable virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f0e04b\": 0.0, \"candidate_e1dca2\": 0.0, \"candidate_009ade\": 0.0, \"candidate_b9213f\": 1.0, \"candidate_822cec\": 0.0, \"candidate_c2dd39\": 0.0, \"candidate_e4165c\": 0.0, \"candidate_fef432\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b367d4cea2ba94c3d7bc3b35:choice", "group_id": "fresh12:banking77:b367d4cea2ba94c3d7bc3b35", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card has been swallowed by an ATM\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2f3960\": \"The user's request concerns lost or stolen card.\", \"candidate_ee54cd\": \"The user's request concerns edit personal details.\", \"candidate_ab5ff2\": \"The user's request concerns card swallowed.\", \"candidate_19b325\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_3d3a19\": \"The user's request concerns why verify identity.\", \"candidate_116904\": \"The user's request concerns top up limits.\", \"candidate_b30c94\": \"The user's request concerns transaction charged twice.\", \"candidate_c0219a\": \"The user's request concerns top up by card charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2f3960\": 0.0, \"candidate_ee54cd\": 0.0, \"candidate_ab5ff2\": 1.0, \"candidate_19b325\": 0.0, \"candidate_3d3a19\": 0.0, \"candidate_116904\": 0.0, \"candidate_b30c94\": 0.0, \"candidate_c0219a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:78f72e0c47313152d5bd8714:choice", "group_id": "fresh12:banking77:78f72e0c47313152d5bd8714", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The ATM won't give back my card\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_69236f\": \"The user's request concerns getting spare card.\", \"candidate_145271\": \"The user's request concerns card swallowed.\", \"candidate_7009d2\": \"The user's request concerns card about to expire.\", \"candidate_b06287\": \"The user's request concerns transfer into account.\", \"candidate_4c23da\": \"The user's request concerns card payment fee charged.\", \"candidate_51c247\": \"The user's request concerns request refund.\", \"candidate_957609\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_4c0947\": \"The user's request concerns card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_69236f\": 0.0, \"candidate_145271\": 1.0, \"candidate_7009d2\": 0.0, \"candidate_b06287\": 0.0, \"candidate_4c23da\": 0.0, \"candidate_51c247\": 0.0, \"candidate_957609\": 0.0, \"candidate_4c0947\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4681397fa2790722e2b15ef6:choice", "group_id": "fresh12:banking77:4681397fa2790722e2b15ef6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The ATM machine stole my card.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a2b343\": \"The user's request concerns top up reverted.\", \"candidate_a586a5\": \"The user's request concerns transfer timing.\", \"candidate_852f9d\": \"The user's request concerns apple pay or google pay.\", \"candidate_76b13f\": \"The user's request concerns atm support.\", \"candidate_52b9e3\": \"The user's request concerns automatic top up.\", \"candidate_657908\": \"The user's request concerns declined cash withdrawal.\", \"candidate_df9af6\": \"The user's request concerns card swallowed.\", \"candidate_adaa78\": \"The user's request concerns pending top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a2b343\": 0.0, \"candidate_a586a5\": 0.0, \"candidate_852f9d\": 0.0, \"candidate_76b13f\": 0.0, \"candidate_52b9e3\": 0.0, \"candidate_657908\": 0.0, \"candidate_df9af6\": 1.0, \"candidate_adaa78\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b30d18a9e27259f2f2ee834f:choice", "group_id": "fresh12:banking77:b30d18a9e27259f2f2ee834f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The ATM didn't give me the card back!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_30129c\": \"The user's request concerns transfer timing.\", \"candidate_e5d900\": \"The user's request concerns pending top up.\", \"candidate_adfcbc\": \"The user's request concerns disposable card limits.\", \"candidate_abd40e\": \"The user's request concerns top up limits.\", \"candidate_796007\": \"The user's request concerns wrong amount of cash received.\", \"candidate_79d9d1\": \"The user's request concerns country support.\", \"candidate_356789\": \"The user's request concerns receiving money.\", \"candidate_135642\": \"The user's request concerns card swallowed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_30129c\": 0.0, \"candidate_e5d900\": 0.0, \"candidate_adfcbc\": 0.0, \"candidate_abd40e\": 0.0, \"candidate_796007\": 0.0, \"candidate_79d9d1\": 0.0, \"candidate_356789\": 0.0, \"candidate_135642\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b73830257e3f1e2fd9a6d71d:choice", "group_id": "fresh12:banking77:b73830257e3f1e2fd9a6d71d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What do I do since the machine swallowed my card. I need one.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0a98d0\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_897a0d\": \"The user's request concerns topping up by card.\", \"candidate_5df22a\": \"The user's request concerns order physical card.\", \"candidate_e15629\": \"The user's request concerns pending top up.\", \"candidate_fe3cb5\": \"The user's request concerns extra charge on statement.\", \"candidate_3b06f1\": \"The user's request concerns declined card payment.\", \"candidate_5d31c0\": \"The user's request concerns card swallowed.\", \"candidate_c58d5b\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0a98d0\": 0.0, \"candidate_897a0d\": 0.0, \"candidate_5df22a\": 0.0, \"candidate_e15629\": 0.0, \"candidate_fe3cb5\": 0.0, \"candidate_3b06f1\": 0.0, \"candidate_5d31c0\": 1.0, \"candidate_c58d5b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a95c1d6510f6221e77438115:choice", "group_id": "fresh12:banking77:a95c1d6510f6221e77438115", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I can't pull my card out of the ATM. Help me.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e09712\": \"The user's request concerns pin blocked.\", \"candidate_901c3a\": \"The user's request concerns card swallowed.\", \"candidate_c2843b\": \"The user's request concerns order physical card.\", \"candidate_c635db\": \"The user's request concerns reverted card payment?.\", \"candidate_9b3137\": \"The user's request concerns compromised card.\", \"candidate_7a1e80\": \"The user's request concerns edit personal details.\", \"candidate_c698b7\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_3d69c6\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e09712\": 0.0, \"candidate_901c3a\": 1.0, \"candidate_c2843b\": 0.0, \"candidate_c635db\": 0.0, \"candidate_9b3137\": 0.0, \"candidate_7a1e80\": 0.0, \"candidate_c698b7\": 0.0, \"candidate_3d69c6\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9aaab92a643756692c366031:choice", "group_id": "fresh12:banking77:9aaab92a643756692c366031", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"It appears that I am being double charged for some items that I have purchased this past week.  Please review and correct.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b50fca\": \"The user's request concerns transaction charged twice.\", \"candidate_237371\": \"The user's request concerns compromised card.\", \"candidate_f2a5fb\": \"The user's request concerns getting spare card.\", \"candidate_a54b77\": \"The user's request concerns transfer fee charged.\", \"candidate_11cfd4\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_be1b06\": \"The user's request concerns why verify identity.\", \"candidate_74b2fc\": \"The user's request concerns get disposable virtual card.\", \"candidate_565a5a\": \"The user's request concerns virtual card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b50fca\": 1.0, \"candidate_237371\": 0.0, \"candidate_f2a5fb\": 0.0, \"candidate_a54b77\": 0.0, \"candidate_11cfd4\": 0.0, \"candidate_be1b06\": 0.0, \"candidate_74b2fc\": 0.0, \"candidate_565a5a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8c06335ff8405aa08256786c:choice", "group_id": "fresh12:banking77:8c06335ff8405aa08256786c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why have I been charged twice for the same transaction?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4d24af\": \"The user's request concerns top up by card charge.\", \"candidate_5d0c8b\": \"The user's request concerns pending cash withdrawal.\", \"candidate_61f3b5\": \"The user's request concerns atm support.\", \"candidate_8ed349\": \"The user's request concerns wrong amount of cash received.\", \"candidate_1f69b3\": \"The user's request concerns transfer fee charged.\", \"candidate_6aaddd\": \"The user's request concerns transaction charged twice.\", \"candidate_30af21\": \"The user's request concerns declined card payment.\", \"candidate_fd689c\": \"The user's request concerns card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4d24af\": 0.0, \"candidate_5d0c8b\": 0.0, \"candidate_61f3b5\": 0.0, \"candidate_8ed349\": 0.0, \"candidate_1f69b3\": 0.0, \"candidate_6aaddd\": 1.0, \"candidate_30af21\": 0.0, \"candidate_fd689c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e987c217e8ba3daf169ce56d:choice", "group_id": "fresh12:banking77:e987c217e8ba3daf169ce56d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Please check payments on my card. There is a duplicate and I only bought it once.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e73f85\": \"The user's request concerns card payment not recognised.\", \"candidate_a63077\": \"The user's request concerns card delivery estimate.\", \"candidate_359dd0\": \"The user's request concerns transaction charged twice.\", \"candidate_039132\": \"The user's request concerns card arrival.\", \"candidate_6e4a6e\": \"The user's request concerns pending cash withdrawal.\", \"candidate_edf669\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_d87027\": \"The user's request concerns top up failed.\", \"candidate_921d30\": \"The user's request concerns automatic top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e73f85\": 0.0, \"candidate_a63077\": 0.0, \"candidate_359dd0\": 1.0, \"candidate_039132\": 0.0, \"candidate_6e4a6e\": 0.0, \"candidate_edf669\": 0.0, \"candidate_d87027\": 0.0, \"candidate_921d30\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1dcfefbd3aa178841aee68e3:choice", "group_id": "fresh12:banking77:1dcfefbd3aa178841aee68e3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I would like to know why I was charged twice for my purchase.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bd19a6\": \"The user's request concerns compromised card.\", \"candidate_89bed4\": \"The user's request concerns Refund not showing up.\", \"candidate_a8bd25\": \"The user's request concerns transaction charged twice.\", \"candidate_c30c6d\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_7b0eda\": \"The user's request concerns get physical card.\", \"candidate_1e8157\": \"The user's request concerns pending card payment.\", \"candidate_b7399e\": \"The user's request concerns card arrival.\", \"candidate_f1e3d9\": \"The user's request concerns card payment wrong exchange rate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bd19a6\": 0.0, \"candidate_89bed4\": 0.0, \"candidate_a8bd25\": 1.0, \"candidate_c30c6d\": 0.0, \"candidate_7b0eda\": 0.0, \"candidate_1e8157\": 0.0, \"candidate_b7399e\": 0.0, \"candidate_f1e3d9\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5b43fb6395145ffd6f4718d2:choice", "group_id": "fresh12:banking77:5b43fb6395145ffd6f4718d2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I see what looks like duplicate charges on account.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7324fd\": \"The user's request concerns transaction charged twice.\", \"candidate_3e6495\": \"The user's request concerns card payment fee charged.\", \"candidate_1d8619\": \"The user's request concerns top up limits.\", \"candidate_fa9017\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_8ef16e\": \"The user's request concerns getting spare card.\", \"candidate_0316e9\": \"The user's request concerns supported cards and currencies.\", \"candidate_bbf7a8\": \"The user's request concerns card acceptance.\", \"candidate_1c9c1f\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7324fd\": 1.0, \"candidate_3e6495\": 0.0, \"candidate_1d8619\": 0.0, \"candidate_fa9017\": 0.0, \"candidate_8ef16e\": 0.0, \"candidate_0316e9\": 0.0, \"candidate_bbf7a8\": 0.0, \"candidate_1c9c1f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:07b67357b90d525b476bc2fd:choice", "group_id": "fresh12:banking77:07b67357b90d525b476bc2fd", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I have been charged twice\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_368fff\": \"The user's request concerns lost or stolen card.\", \"candidate_5ce8d3\": \"The user's request concerns declined card payment.\", \"candidate_3cb0cd\": \"The user's request concerns transfer into account.\", \"candidate_429b6a\": \"The user's request concerns card linking.\", \"candidate_1b1b08\": \"The user's request concerns transfer timing.\", \"candidate_40e7e6\": \"The user's request concerns wrong amount of cash received.\", \"candidate_01d224\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_a1b19f\": \"The user's request concerns transaction charged twice.\"}}}}", "targets_json": "{\"decision\": {\"candidate_368fff\": 0.0, \"candidate_5ce8d3\": 0.0, \"candidate_3cb0cd\": 0.0, \"candidate_429b6a\": 0.0, \"candidate_1b1b08\": 0.0, \"candidate_40e7e6\": 0.0, \"candidate_01d224\": 0.0, \"candidate_a1b19f\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e2108abf6e77ff173ef7b216:choice", "group_id": "fresh12:banking77:e2108abf6e77ff173ef7b216", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can I see my source of funds?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5df556\": \"The user's request concerns automatic top up.\", \"candidate_2e8f1f\": \"The user's request concerns top up failed.\", \"candidate_bf0296\": \"The user's request concerns getting virtual card.\", \"candidate_9730c2\": \"The user's request concerns verify source of funds.\", \"candidate_58f414\": \"The user's request concerns lost or stolen phone.\", \"candidate_864c90\": \"The user's request concerns topping up by card.\", \"candidate_a009eb\": \"The user's request concerns card about to expire.\", \"candidate_92194b\": \"The user's request concerns card acceptance.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5df556\": 0.0, \"candidate_2e8f1f\": 0.0, \"candidate_bf0296\": 0.0, \"candidate_9730c2\": 1.0, \"candidate_58f414\": 0.0, \"candidate_864c90\": 0.0, \"candidate_a009eb\": 0.0, \"candidate_92194b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:31438646fba09f442cb2ce3d:choice", "group_id": "fresh12:banking77:31438646fba09f442cb2ce3d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can I tell where my funds came from?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bcb230\": \"The user's request concerns exchange via app.\", \"candidate_f7729f\": \"The user's request concerns cash withdrawal charge.\", \"candidate_89440c\": \"The user's request concerns fiat currency support.\", \"candidate_e5b929\": \"The user's request concerns Refund not showing up.\", \"candidate_077b66\": \"The user's request concerns extra charge on statement.\", \"candidate_1454a1\": \"The user's request concerns supported cards and currencies.\", \"candidate_ca658b\": \"The user's request concerns verify source of funds.\", \"candidate_235c3d\": \"The user's request concerns wrong amount of cash received.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bcb230\": 0.0, \"candidate_f7729f\": 0.0, \"candidate_89440c\": 0.0, \"candidate_e5b929\": 0.0, \"candidate_077b66\": 0.0, \"candidate_1454a1\": 0.0, \"candidate_ca658b\": 1.0, \"candidate_235c3d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4f4514b37ddb6a6fc3aaa7fd:choice", "group_id": "fresh12:banking77:4f4514b37ddb6a6fc3aaa7fd", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What is the source of my funds.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_79a600\": \"The user's request concerns declined transfer.\", \"candidate_a081c3\": \"The user's request concerns getting virtual card.\", \"candidate_714796\": \"The user's request concerns transaction charged twice.\", \"candidate_7876a4\": \"The user's request concerns extra charge on statement.\", \"candidate_e93294\": \"The user's request concerns get physical card.\", \"candidate_989283\": \"The user's request concerns apple pay or google pay.\", \"candidate_6be323\": \"The user's request concerns verify source of funds.\", \"candidate_e11e6e\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_79a600\": 0.0, \"candidate_a081c3\": 0.0, \"candidate_714796\": 0.0, \"candidate_7876a4\": 0.0, \"candidate_e93294\": 0.0, \"candidate_989283\": 0.0, \"candidate_6be323\": 1.0, \"candidate_e11e6e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4f2ad039c24db62aa770636c:choice", "group_id": "fresh12:banking77:4f2ad039c24db62aa770636c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where are my funds coming from? I need to know.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_85f403\": \"The user's request concerns verify source of funds.\", \"candidate_ce78e4\": \"The user's request concerns visa or mastercard.\", \"candidate_bc2468\": \"The user's request concerns transfer timing.\", \"candidate_224fcc\": \"The user's request concerns beneficiary not allowed.\", \"candidate_addb77\": \"The user's request concerns reverted card payment?.\", \"candidate_15dcc7\": \"The user's request concerns card payment not recognised.\", \"candidate_9f80f3\": \"The user's request concerns pending cash withdrawal.\", \"candidate_6c99eb\": \"The user's request concerns declined card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_85f403\": 1.0, \"candidate_ce78e4\": 0.0, \"candidate_bc2468\": 0.0, \"candidate_224fcc\": 0.0, \"candidate_addb77\": 0.0, \"candidate_15dcc7\": 0.0, \"candidate_9f80f3\": 0.0, \"candidate_6c99eb\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:171e077e65925b2427105027:choice", "group_id": "fresh12:banking77:171e077e65925b2427105027", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"where to funds come from\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9e39e5\": \"The user's request concerns pending cash withdrawal.\", \"candidate_bd2f44\": \"The user's request concerns top up reverted.\", \"candidate_dde3d2\": \"The user's request concerns verify top up.\", \"candidate_1593da\": \"The user's request concerns unable to verify identity.\", \"candidate_d6cbf2\": \"The user's request concerns pin blocked.\", \"candidate_f54bea\": \"The user's request concerns disposable card limits.\", \"candidate_36813b\": \"The user's request concerns card payment not recognised.\", \"candidate_03ffdc\": \"The user's request concerns verify source of funds.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9e39e5\": 0.0, \"candidate_bd2f44\": 0.0, \"candidate_dde3d2\": 0.0, \"candidate_1593da\": 0.0, \"candidate_d6cbf2\": 0.0, \"candidate_f54bea\": 0.0, \"candidate_36813b\": 0.0, \"candidate_03ffdc\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:237053bfd8f5e1155ce95699:choice", "group_id": "fresh12:banking77:237053bfd8f5e1155ce95699", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I would like to see the source of my money.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a09741\": \"The user's request concerns why verify identity.\", \"candidate_caa841\": \"The user's request concerns verify source of funds.\", \"candidate_9a3dc4\": \"The user's request concerns top up failed.\", \"candidate_c105e8\": \"The user's request concerns receiving money.\", \"candidate_288e24\": \"The user's request concerns transfer timing.\", \"candidate_94ec67\": \"The user's request concerns exchange charge.\", \"candidate_b58b3c\": \"The user's request concerns verify top up.\", \"candidate_9c6ac4\": \"The user's request concerns top up by card charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a09741\": 0.0, \"candidate_caa841\": 1.0, \"candidate_9a3dc4\": 0.0, \"candidate_c105e8\": 0.0, \"candidate_288e24\": 0.0, \"candidate_94ec67\": 0.0, \"candidate_b58b3c\": 0.0, \"candidate_9c6ac4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d3a9e623e1719f31f46a4d96:choice", "group_id": "fresh12:banking77:d3a9e623e1719f31f46a4d96", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"When does the money get transferred to my account\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9d372d\": \"The user's request concerns failed transfer.\", \"candidate_59dfae\": \"The user's request concerns card payment not recognised.\", \"candidate_20a378\": \"The user's request concerns exchange via app.\", \"candidate_4c883d\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_886510\": \"The user's request concerns transfer timing.\", \"candidate_d56ecb\": \"The user's request concerns getting virtual card.\", \"candidate_9ca3ec\": \"The user's request concerns top up failed.\", \"candidate_5a0188\": \"The user's request concerns exchange charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9d372d\": 0.0, \"candidate_59dfae\": 0.0, \"candidate_20a378\": 0.0, \"candidate_4c883d\": 0.0, \"candidate_886510\": 1.0, \"candidate_d56ecb\": 0.0, \"candidate_9ca3ec\": 0.0, \"candidate_5a0188\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:87fd72e8452369cff7a35dc7:choice", "group_id": "fresh12:banking77:87fd72e8452369cff7a35dc7", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long will it take for the transfer to reach my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2af641\": \"The user's request concerns card payment fee charged.\", \"candidate_8fdd8f\": \"The user's request concerns reverted card payment?.\", \"candidate_22ed06\": \"The user's request concerns unable to verify identity.\", \"candidate_5a1981\": \"The user's request concerns transfer timing.\", \"candidate_fac01a\": \"The user's request concerns declined card payment.\", \"candidate_e204e2\": \"The user's request concerns top up by card charge.\", \"candidate_9bb4db\": \"The user's request concerns edit personal details.\", \"candidate_fe26f5\": \"The user's request concerns card payment wrong exchange rate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2af641\": 0.0, \"candidate_8fdd8f\": 0.0, \"candidate_22ed06\": 0.0, \"candidate_5a1981\": 1.0, \"candidate_fac01a\": 0.0, \"candidate_e204e2\": 0.0, \"candidate_9bb4db\": 0.0, \"candidate_fe26f5\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2f845d17e574cabea0c036d4:choice", "group_id": "fresh12:banking77:2f845d17e574cabea0c036d4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How many days does a transfer from the US take?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d3f6fb\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_84de1d\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_6071d5\": \"The user's request concerns card acceptance.\", \"candidate_f069c8\": \"The user's request concerns transfer timing.\", \"candidate_a964ca\": \"The user's request concerns getting spare card.\", \"candidate_7a5f9c\": \"The user's request concerns top up reverted.\", \"candidate_7bbb03\": \"The user's request concerns why verify identity.\", \"candidate_6a320e\": \"The user's request concerns compromised card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d3f6fb\": 0.0, \"candidate_84de1d\": 0.0, \"candidate_6071d5\": 0.0, \"candidate_f069c8\": 1.0, \"candidate_a964ca\": 0.0, \"candidate_7a5f9c\": 0.0, \"candidate_7bbb03\": 0.0, \"candidate_6a320e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:db771e9ad3771a9084aa093b:choice", "group_id": "fresh12:banking77:db771e9ad3771a9084aa093b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long until the money is in my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4bd7cd\": \"The user's request concerns wrong amount of cash received.\", \"candidate_787067\": \"The user's request concerns top up by card charge.\", \"candidate_fa4659\": \"The user's request concerns terminate account.\", \"candidate_1cebbf\": \"The user's request concerns card acceptance.\", \"candidate_c16165\": \"The user's request concerns supported cards and currencies.\", \"candidate_178702\": \"The user's request concerns pending cash withdrawal.\", \"candidate_40d882\": \"The user's request concerns age limit.\", \"candidate_719ade\": \"The user's request concerns transfer timing.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4bd7cd\": 0.0, \"candidate_787067\": 0.0, \"candidate_fa4659\": 0.0, \"candidate_1cebbf\": 0.0, \"candidate_c16165\": 0.0, \"candidate_178702\": 0.0, \"candidate_40d882\": 0.0, \"candidate_719ade\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f7c9490d99a9565f3659acdb:choice", "group_id": "fresh12:banking77:f7c9490d99a9565f3659acdb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"when will i receive a US transfer\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ab08bd\": \"The user's request concerns terminate account.\", \"candidate_7bf648\": \"The user's request concerns transfer timing.\", \"candidate_6603f8\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_b42e77\": \"The user's request concerns pending cash withdrawal.\", \"candidate_df58e7\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_7864b6\": \"The user's request concerns pin blocked.\", \"candidate_3e7347\": \"The user's request concerns fiat currency support.\", \"candidate_1a996d\": \"The user's request concerns lost or stolen phone.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ab08bd\": 0.0, \"candidate_7bf648\": 1.0, \"candidate_6603f8\": 0.0, \"candidate_b42e77\": 0.0, \"candidate_df58e7\": 0.0, \"candidate_7864b6\": 0.0, \"candidate_3e7347\": 0.0, \"candidate_1a996d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:054ef9d5dec27f420d61ecea:choice", "group_id": "fresh12:banking77:054ef9d5dec27f420d61ecea", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long does a transfer take from the point it's sent, to the point it arrives in my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_03fe85\": \"The user's request concerns transfer timing.\", \"candidate_44c74b\": \"The user's request concerns top up reverted.\", \"candidate_6303bd\": \"The user's request concerns verify source of funds.\", \"candidate_3c1db3\": \"The user's request concerns failed transfer.\", \"candidate_e7d8d4\": \"The user's request concerns cancel transfer.\", \"candidate_adc8e8\": \"The user's request concerns pending cash withdrawal.\", \"candidate_26dd7f\": \"The user's request concerns change pin.\", \"candidate_892731\": \"The user's request concerns terminate account.\"}}}}", "targets_json": "{\"decision\": {\"candidate_03fe85\": 1.0, \"candidate_44c74b\": 0.0, \"candidate_6303bd\": 0.0, \"candidate_3c1db3\": 0.0, \"candidate_e7d8d4\": 0.0, \"candidate_adc8e8\": 0.0, \"candidate_26dd7f\": 0.0, \"candidate_892731\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c0ad5c9ab4f687cfcf79e69a:choice", "group_id": "fresh12:banking77:c0ad5c9ab4f687cfcf79e69a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Someone stopped my payment\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d5cef4\": \"The user's request concerns why verify identity.\", \"candidate_08d488\": \"The user's request concerns card payment not recognised.\", \"candidate_cb65df\": \"The user's request concerns declined card payment.\", \"candidate_92da6f\": \"The user's request concerns Refund not showing up.\", \"candidate_066ed1\": \"The user's request concerns reverted card payment?.\", \"candidate_3d5fe8\": \"The user's request concerns pending transfer.\", \"candidate_9a1506\": \"The user's request concerns card payment fee charged.\", \"candidate_51c302\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d5cef4\": 0.0, \"candidate_08d488\": 0.0, \"candidate_cb65df\": 0.0, \"candidate_92da6f\": 0.0, \"candidate_066ed1\": 1.0, \"candidate_3d5fe8\": 0.0, \"candidate_9a1506\": 0.0, \"candidate_51c302\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:70d9a7ba7c7a62fe7712aa3a:choice", "group_id": "fresh12:banking77:70d9a7ba7c7a62fe7712aa3a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card is being declined for a purchase. I bought items before and the card worked. Do you know what the problem is?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c5b189\": \"The user's request concerns fiat currency support.\", \"candidate_78d526\": \"The user's request concerns reverted card payment?.\", \"candidate_6dce5b\": \"The user's request concerns verify source of funds.\", \"candidate_823ed7\": \"The user's request concerns declined card payment.\", \"candidate_544733\": \"The user's request concerns edit personal details.\", \"candidate_9aa805\": \"The user's request concerns card arrival.\", \"candidate_c7c33e\": \"The user's request concerns activate my card.\", \"candidate_60f8d2\": \"The user's request concerns getting spare card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c5b189\": 0.0, \"candidate_78d526\": 1.0, \"candidate_6dce5b\": 0.0, \"candidate_823ed7\": 0.0, \"candidate_544733\": 0.0, \"candidate_9aa805\": 0.0, \"candidate_c7c33e\": 0.0, \"candidate_60f8d2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:283888769f9a52d5492a6984:choice", "group_id": "fresh12:banking77:283888769f9a52d5492a6984", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I bought something and the money appeared back into my account? Why?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_736f07\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_5930a9\": \"The user's request concerns change pin.\", \"candidate_d39e93\": \"The user's request concerns transfer into account.\", \"candidate_a0121e\": \"The user's request concerns extra charge on statement.\", \"candidate_78654d\": \"The user's request concerns get disposable virtual card.\", \"candidate_8b3b51\": \"The user's request concerns card payment not recognised.\", \"candidate_e3b31b\": \"The user's request concerns transfer timing.\", \"candidate_f19951\": \"The user's request concerns reverted card payment?.\"}}}}", "targets_json": "{\"decision\": {\"candidate_736f07\": 0.0, \"candidate_5930a9\": 0.0, \"candidate_d39e93\": 0.0, \"candidate_a0121e\": 0.0, \"candidate_78654d\": 0.0, \"candidate_8b3b51\": 0.0, \"candidate_e3b31b\": 0.0, \"candidate_f19951\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:30a0b28d974474ce21988fa2:choice", "group_id": "fresh12:banking77:30a0b28d974474ce21988fa2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I think my card payment has been stopped\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_31e0aa\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_75af8f\": \"The user's request concerns top up by cash or cheque.\", \"candidate_370e71\": \"The user's request concerns request refund.\", \"candidate_5eda45\": \"The user's request concerns passcode forgotten.\", \"candidate_9118f3\": \"The user's request concerns reverted card payment?.\", \"candidate_3ade99\": \"The user's request concerns card payment fee charged.\", \"candidate_097ef2\": \"The user's request concerns cash withdrawal charge.\", \"candidate_964ea6\": \"The user's request concerns verify my identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_31e0aa\": 0.0, \"candidate_75af8f\": 0.0, \"candidate_370e71\": 0.0, \"candidate_5eda45\": 0.0, \"candidate_9118f3\": 1.0, \"candidate_3ade99\": 0.0, \"candidate_097ef2\": 0.0, \"candidate_964ea6\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:212cf77eb78a14ac4019404a:choice", "group_id": "fresh12:banking77:212cf77eb78a14ac4019404a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why was my card payment cancelled?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1dee94\": \"The user's request concerns reverted card payment?.\", \"candidate_70ebfd\": \"The user's request concerns transfer fee charged.\", \"candidate_5a0fed\": \"The user's request concerns cancel transfer.\", \"candidate_a745da\": \"The user's request concerns card payment fee charged.\", \"candidate_18e648\": \"The user's request concerns edit personal details.\", \"candidate_b3fbb9\": \"The user's request concerns verify top up.\", \"candidate_f9c7d7\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_e02eb2\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1dee94\": 1.0, \"candidate_70ebfd\": 0.0, \"candidate_5a0fed\": 0.0, \"candidate_a745da\": 0.0, \"candidate_18e648\": 0.0, \"candidate_b3fbb9\": 0.0, \"candidate_f9c7d7\": 0.0, \"candidate_e02eb2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ef3c3db015000db4d35d9be8:choice", "group_id": "fresh12:banking77:ef3c3db015000db4d35d9be8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I did a payment but the app reverted it\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d52ec8\": \"The user's request concerns atm support.\", \"candidate_b6fdb3\": \"The user's request concerns reverted card payment?.\", \"candidate_97dc0d\": \"The user's request concerns top up by card charge.\", \"candidate_60b41f\": \"The user's request concerns visa or mastercard.\", \"candidate_9e9e7c\": \"The user's request concerns exchange via app.\", \"candidate_fe59af\": \"The user's request concerns card swallowed.\", \"candidate_249f6e\": \"The user's request concerns order physical card.\", \"candidate_e71d11\": \"The user's request concerns beneficiary not allowed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d52ec8\": 0.0, \"candidate_b6fdb3\": 1.0, \"candidate_97dc0d\": 0.0, \"candidate_60b41f\": 0.0, \"candidate_9e9e7c\": 0.0, \"candidate_fe59af\": 0.0, \"candidate_249f6e\": 0.0, \"candidate_e71d11\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a361a9d406476db81d83a6e9:choice", "group_id": "fresh12:banking77:a361a9d406476db81d83a6e9", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I am in Austria right now. I need to change my PIN ASAP, so can I still do this from here?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dd2103\": \"The user's request concerns change pin.\", \"candidate_20bb3a\": \"The user's request concerns top up by card charge.\", \"candidate_713452\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_4aa25a\": \"The user's request concerns pending cash withdrawal.\", \"candidate_b69678\": \"The user's request concerns card payment not recognised.\", \"candidate_74d7ae\": \"The user's request concerns top up reverted.\", \"candidate_7a95be\": \"The user's request concerns pin blocked.\", \"candidate_4b4f73\": \"The user's request concerns age limit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dd2103\": 1.0, \"candidate_20bb3a\": 0.0, \"candidate_713452\": 0.0, \"candidate_4aa25a\": 0.0, \"candidate_b69678\": 0.0, \"candidate_74d7ae\": 0.0, \"candidate_7a95be\": 0.0, \"candidate_4b4f73\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2c3c0bad64ea30d885069ac3:choice", "group_id": "fresh12:banking77:2c3c0bad64ea30d885069ac3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where do I change my PIN?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_91891b\": \"The user's request concerns edit personal details.\", \"candidate_169de5\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_c51bbc\": \"The user's request concerns change pin.\", \"candidate_bad4c0\": \"The user's request concerns activate my card.\", \"candidate_45e8eb\": \"The user's request concerns transfer timing.\", \"candidate_007c4c\": \"The user's request concerns reverted card payment?.\", \"candidate_e56046\": \"The user's request concerns top up by card charge.\", \"candidate_2c8dbc\": \"The user's request concerns card swallowed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_91891b\": 0.0, \"candidate_169de5\": 0.0, \"candidate_c51bbc\": 1.0, \"candidate_bad4c0\": 0.0, \"candidate_45e8eb\": 0.0, \"candidate_007c4c\": 0.0, \"candidate_e56046\": 0.0, \"candidate_2c8dbc\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:892dfb4efdb5e0eaa9aaed69:choice", "group_id": "fresh12:banking77:892dfb4efdb5e0eaa9aaed69", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I want to choose a different PIN.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_320bdc\": \"The user's request concerns contactless not working.\", \"candidate_a583d5\": \"The user's request concerns disposable card limits.\", \"candidate_fd983e\": \"The user's request concerns exchange charge.\", \"candidate_9dba5d\": \"The user's request concerns card arrival.\", \"candidate_59f831\": \"The user's request concerns unable to verify identity.\", \"candidate_dec7ae\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_5e1eaa\": \"The user's request concerns change pin.\", \"candidate_170f03\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_320bdc\": 0.0, \"candidate_a583d5\": 0.0, \"candidate_fd983e\": 0.0, \"candidate_9dba5d\": 0.0, \"candidate_59f831\": 0.0, \"candidate_dec7ae\": 0.0, \"candidate_5e1eaa\": 1.0, \"candidate_170f03\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0178adcd308aa372349169f4:choice", "group_id": "fresh12:banking77:0178adcd308aa372349169f4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need a new PIN.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ad2779\": \"The user's request concerns exchange rate.\", \"candidate_ed9612\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_85d44e\": \"The user's request concerns order physical card.\", \"candidate_72c369\": \"The user's request concerns pending top up.\", \"candidate_09740e\": \"The user's request concerns transaction charged twice.\", \"candidate_161830\": \"The user's request concerns change pin.\", \"candidate_a5107b\": \"The user's request concerns extra charge on statement.\", \"candidate_92801d\": \"The user's request concerns transfer timing.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ad2779\": 0.0, \"candidate_ed9612\": 0.0, \"candidate_85d44e\": 0.0, \"candidate_72c369\": 0.0, \"candidate_09740e\": 0.0, \"candidate_161830\": 1.0, \"candidate_a5107b\": 0.0, \"candidate_92801d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e57dfb7e7b9aa4d1cfb4e456:choice", "group_id": "fresh12:banking77:e57dfb7e7b9aa4d1cfb4e456", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Help me set up a new PIN?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_885995\": \"The user's request concerns change pin.\", \"candidate_3a0620\": \"The user's request concerns transfer fee charged.\", \"candidate_d43c02\": \"The user's request concerns reverted card payment?.\", \"candidate_18ffd5\": \"The user's request concerns visa or mastercard.\", \"candidate_64bfb1\": \"The user's request concerns card about to expire.\", \"candidate_5bbb0a\": \"The user's request concerns terminate account.\", \"candidate_f5ce5b\": \"The user's request concerns disposable card limits.\", \"candidate_79b6de\": \"The user's request concerns getting spare card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_885995\": 1.0, \"candidate_3a0620\": 0.0, \"candidate_d43c02\": 0.0, \"candidate_18ffd5\": 0.0, \"candidate_64bfb1\": 0.0, \"candidate_5bbb0a\": 0.0, \"candidate_f5ce5b\": 0.0, \"candidate_79b6de\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5f6d4af859b45669eca61f0b:choice", "group_id": "fresh12:banking77:5f6d4af859b45669eca61f0b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a location near me that i can change my PIN?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9b983d\": \"The user's request concerns failed transfer.\", \"candidate_4d4fee\": \"The user's request concerns receiving money.\", \"candidate_cb4058\": \"The user's request concerns verify my identity.\", \"candidate_59f84b\": \"The user's request concerns activate my card.\", \"candidate_443214\": \"The user's request concerns change pin.\", \"candidate_c9e72c\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_98dfcb\": \"The user's request concerns card linking.\", \"candidate_7c9e6e\": \"The user's request concerns virtual card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9b983d\": 0.0, \"candidate_4d4fee\": 0.0, \"candidate_cb4058\": 0.0, \"candidate_59f84b\": 0.0, \"candidate_443214\": 1.0, \"candidate_c9e72c\": 0.0, \"candidate_98dfcb\": 0.0, \"candidate_7c9e6e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6d8ea75be2e3bab52d88e9e5:choice", "group_id": "fresh12:banking77:6d8ea75be2e3bab52d88e9e5", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I transfer money to a beneficiery?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_575c0d\": \"The user's request concerns lost or stolen card.\", \"candidate_fd0a4a\": \"The user's request concerns pending cash withdrawal.\", \"candidate_fc88fd\": \"The user's request concerns Refund not showing up.\", \"candidate_ec832d\": \"The user's request concerns visa or mastercard.\", \"candidate_9d1add\": \"The user's request concerns pin blocked.\", \"candidate_389e30\": \"The user's request concerns beneficiary not allowed.\", \"candidate_085208\": \"The user's request concerns automatic top up.\", \"candidate_a5faaa\": \"The user's request concerns topping up by card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_575c0d\": 0.0, \"candidate_fd0a4a\": 0.0, \"candidate_fc88fd\": 0.0, \"candidate_ec832d\": 0.0, \"candidate_9d1add\": 0.0, \"candidate_389e30\": 1.0, \"candidate_085208\": 0.0, \"candidate_a5faaa\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d2e0ea818add4026184202d1:choice", "group_id": "fresh12:banking77:d2e0ea818add4026184202d1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What's the deal with no cryptocurrency on your app?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5d385e\": \"The user's request concerns verify top up.\", \"candidate_0f09e3\": \"The user's request concerns apple pay or google pay.\", \"candidate_48fa4e\": \"The user's request concerns visa or mastercard.\", \"candidate_a4201e\": \"The user's request concerns unable to verify identity.\", \"candidate_b17224\": \"The user's request concerns beneficiary not allowed.\", \"candidate_ca38d1\": \"The user's request concerns card about to expire.\", \"candidate_75401a\": \"The user's request concerns lost or stolen phone.\", \"candidate_f46eee\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5d385e\": 0.0, \"candidate_0f09e3\": 0.0, \"candidate_48fa4e\": 0.0, \"candidate_a4201e\": 0.0, \"candidate_b17224\": 1.0, \"candidate_ca38d1\": 0.0, \"candidate_75401a\": 0.0, \"candidate_f46eee\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:10f753bac3a4caa6eb2626ea:choice", "group_id": "fresh12:banking77:10f753bac3a4caa6eb2626ea", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"why isnt my account allowing transfers\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9f91a1\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_c10d90\": \"The user's request concerns declined card payment.\", \"candidate_0f57e1\": \"The user's request concerns top up failed.\", \"candidate_ca3ae8\": \"The user's request concerns wrong amount of cash received.\", \"candidate_d08344\": \"The user's request concerns beneficiary not allowed.\", \"candidate_a400e0\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_60f9a0\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_e84ec7\": \"The user's request concerns lost or stolen card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9f91a1\": 0.0, \"candidate_c10d90\": 0.0, \"candidate_0f57e1\": 0.0, \"candidate_ca3ae8\": 0.0, \"candidate_d08344\": 1.0, \"candidate_a400e0\": 0.0, \"candidate_60f9a0\": 0.0, \"candidate_e84ec7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:88567c20a2f4fc50ee76ea79:choice", "group_id": "fresh12:banking77:88567c20a2f4fc50ee76ea79", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I want to buy cryptocurrency, but your app doesn't let me. Why?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4e330f\": \"The user's request concerns pin blocked.\", \"candidate_404bb7\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_6b323b\": \"The user's request concerns beneficiary not allowed.\", \"candidate_002bf0\": \"The user's request concerns card delivery estimate.\", \"candidate_c34b6c\": \"The user's request concerns age limit.\", \"candidate_05aa06\": \"The user's request concerns top up failed.\", \"candidate_522493\": \"The user's request concerns verify my identity.\", \"candidate_caa00e\": \"The user's request concerns verify source of funds.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4e330f\": 0.0, \"candidate_404bb7\": 0.0, \"candidate_6b323b\": 1.0, \"candidate_002bf0\": 0.0, \"candidate_c34b6c\": 0.0, \"candidate_05aa06\": 0.0, \"candidate_522493\": 0.0, \"candidate_caa00e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:627e7d5005f8567844e3a53a:choice", "group_id": "fresh12:banking77:627e7d5005f8567844e3a53a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I couldn't make a transfer to a beneficiary!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2e8f2b\": \"The user's request concerns get disposable virtual card.\", \"candidate_6b77de\": \"The user's request concerns passcode forgotten.\", \"candidate_dbe17c\": \"The user's request concerns beneficiary not allowed.\", \"candidate_575f9d\": \"The user's request concerns verify my identity.\", \"candidate_11f396\": \"The user's request concerns exchange rate.\", \"candidate_aada1f\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_086bc5\": \"The user's request concerns lost or stolen card.\", \"candidate_fd3420\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2e8f2b\": 0.0, \"candidate_6b77de\": 0.0, \"candidate_dbe17c\": 1.0, \"candidate_575f9d\": 0.0, \"candidate_11f396\": 0.0, \"candidate_aada1f\": 0.0, \"candidate_086bc5\": 0.0, \"candidate_fd3420\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9b56e9b63c28884eb924c6c9:choice", "group_id": "fresh12:banking77:9b56e9b63c28884eb924c6c9", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I tried to transfer money but it said it wasn't possible? I've done this before and it worked why isn't it working now?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_28dcb9\": \"The user's request concerns get physical card.\", \"candidate_74f866\": \"The user's request concerns beneficiary not allowed.\", \"candidate_89b732\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_3b7ae8\": \"The user's request concerns card swallowed.\", \"candidate_30e386\": \"The user's request concerns declined cash withdrawal.\", \"candidate_d0f8c7\": \"The user's request concerns pending transfer.\", \"candidate_fb718f\": \"The user's request concerns wrong amount of cash received.\", \"candidate_da4075\": \"The user's request concerns top up limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_28dcb9\": 0.0, \"candidate_74f866\": 1.0, \"candidate_89b732\": 0.0, \"candidate_3b7ae8\": 0.0, \"candidate_30e386\": 0.0, \"candidate_d0f8c7\": 0.0, \"candidate_fb718f\": 0.0, \"candidate_da4075\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0e0ddc8840b1e22c64319a79:choice", "group_id": "fresh12:banking77:0e0ddc8840b1e22c64319a79", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"In the past month I often used this account to manage payments for my new holiday villa. Now all of a sudden your fees have increased exorbitantly! Don't you have a rewards programme for frequent users?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bdf656\": \"The user's request concerns wrong amount of cash received.\", \"candidate_628a0a\": \"The user's request concerns transfer fee charged.\", \"candidate_dfb64d\": \"The user's request concerns transfer timing.\", \"candidate_4642c1\": \"The user's request concerns top up by card charge.\", \"candidate_e4274c\": \"The user's request concerns top up failed.\", \"candidate_424815\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_ad5702\": \"The user's request concerns card arrival.\", \"candidate_6a351c\": \"The user's request concerns fiat currency support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bdf656\": 0.0, \"candidate_628a0a\": 1.0, \"candidate_dfb64d\": 0.0, \"candidate_4642c1\": 0.0, \"candidate_e4274c\": 0.0, \"candidate_424815\": 0.0, \"candidate_ad5702\": 0.0, \"candidate_6a351c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:bc5852d22d2a0669daaf1e86:choice", "group_id": "fresh12:banking77:bc5852d22d2a0669daaf1e86", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I was charge a fee for my transfer, why?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_50e14a\": \"The user's request concerns request refund.\", \"candidate_776b07\": \"The user's request concerns exchange rate.\", \"candidate_ee49b8\": \"The user's request concerns transfer fee charged.\", \"candidate_8ba550\": \"The user's request concerns atm support.\", \"candidate_535010\": \"The user's request concerns transfer not received by recipient.\", \"candidate_e67abf\": \"The user's request concerns country support.\", \"candidate_497e5c\": \"The user's request concerns transaction charged twice.\", \"candidate_fc811b\": \"The user's request concerns receiving money.\"}}}}", "targets_json": "{\"decision\": {\"candidate_50e14a\": 0.0, \"candidate_776b07\": 0.0, \"candidate_ee49b8\": 1.0, \"candidate_8ba550\": 0.0, \"candidate_535010\": 0.0, \"candidate_e67abf\": 0.0, \"candidate_497e5c\": 0.0, \"candidate_fc811b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7124cdb15e3f5389c5c33602:choice", "group_id": "fresh12:banking77:7124cdb15e3f5389c5c33602", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can you explain the transfer fee to me?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c32550\": \"The user's request concerns transfer into account.\", \"candidate_f83435\": \"The user's request concerns topping up by card.\", \"candidate_89b9db\": \"The user's request concerns top up limits.\", \"candidate_d21eb6\": \"The user's request concerns receiving money.\", \"candidate_dd9a5d\": \"The user's request concerns transfer fee charged.\", \"candidate_3493c9\": \"The user's request concerns virtual card not working.\", \"candidate_8dffe2\": \"The user's request concerns atm support.\", \"candidate_db7712\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c32550\": 0.0, \"candidate_f83435\": 0.0, \"candidate_89b9db\": 0.0, \"candidate_d21eb6\": 0.0, \"candidate_dd9a5d\": 1.0, \"candidate_3493c9\": 0.0, \"candidate_8dffe2\": 0.0, \"candidate_db7712\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:be3f9341d4042df91fb5dd7b:choice", "group_id": "fresh12:banking77:be3f9341d4042df91fb5dd7b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there supposed to be a fee for the transfer I made?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9ace7d\": \"The user's request concerns declined card payment.\", \"candidate_865927\": \"The user's request concerns fiat currency support.\", \"candidate_3ae373\": \"The user's request concerns contactless not working.\", \"candidate_a00ddd\": \"The user's request concerns why verify identity.\", \"candidate_c7020c\": \"The user's request concerns top up by cash or cheque.\", \"candidate_e02d7d\": \"The user's request concerns transfer fee charged.\", \"candidate_d31335\": \"The user's request concerns card delivery estimate.\", \"candidate_8e56c3\": \"The user's request concerns disposable card limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9ace7d\": 0.0, \"candidate_865927\": 0.0, \"candidate_3ae373\": 0.0, \"candidate_a00ddd\": 0.0, \"candidate_c7020c\": 0.0, \"candidate_e02d7d\": 1.0, \"candidate_d31335\": 0.0, \"candidate_8e56c3\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5820bb7a488a1eb222e00760:choice", "group_id": "fresh12:banking77:5820bb7a488a1eb222e00760", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I sent out money, but was charged extra for transferring. Why?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9e07d3\": \"The user's request concerns Refund not showing up.\", \"candidate_ae8de8\": \"The user's request concerns getting virtual card.\", \"candidate_8440a2\": \"The user's request concerns top up limits.\", \"candidate_f14897\": \"The user's request concerns cancel transfer.\", \"candidate_6de5f5\": \"The user's request concerns transfer not received by recipient.\", \"candidate_8e76cb\": \"The user's request concerns unable to verify identity.\", \"candidate_c1ea44\": \"The user's request concerns transfer fee charged.\", \"candidate_52735b\": \"The user's request concerns balance not updated after cheque or cash deposit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9e07d3\": 0.0, \"candidate_ae8de8\": 0.0, \"candidate_8440a2\": 0.0, \"candidate_f14897\": 0.0, \"candidate_6de5f5\": 0.0, \"candidate_8e76cb\": 0.0, \"candidate_c1ea44\": 1.0, \"candidate_52735b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7d00e5f3ee902453b5994921:choice", "group_id": "fresh12:banking77:7d00e5f3ee902453b5994921", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a reason I was charged a fee to transfer?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a5e7e7\": \"The user's request concerns beneficiary not allowed.\", \"candidate_69c2a7\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_10de52\": \"The user's request concerns pin blocked.\", \"candidate_2174c6\": \"The user's request concerns pending transfer.\", \"candidate_c59554\": \"The user's request concerns change pin.\", \"candidate_9e6b56\": \"The user's request concerns request refund.\", \"candidate_016519\": \"The user's request concerns transfer fee charged.\", \"candidate_392f51\": \"The user's request concerns Refund not showing up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a5e7e7\": 0.0, \"candidate_69c2a7\": 0.0, \"candidate_10de52\": 0.0, \"candidate_2174c6\": 0.0, \"candidate_c59554\": 0.0, \"candidate_9e6b56\": 0.0, \"candidate_016519\": 1.0, \"candidate_392f51\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:16e7d02deb0b2ddb26f23046:choice", "group_id": "fresh12:banking77:16e7d02deb0b2ddb26f23046", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I get paid in GBP. Should I configure this and if so, where?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2554a3\": \"The user's request concerns unable to verify identity.\", \"candidate_5ab23a\": \"The user's request concerns card delivery estimate.\", \"candidate_f75974\": \"The user's request concerns receiving money.\", \"candidate_ee3275\": \"The user's request concerns verify top up.\", \"candidate_ced601\": \"The user's request concerns cancel transfer.\", \"candidate_8e392d\": \"The user's request concerns lost or stolen phone.\", \"candidate_bb491a\": \"The user's request concerns supported cards and currencies.\", \"candidate_e558d6\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2554a3\": 0.0, \"candidate_5ab23a\": 0.0, \"candidate_f75974\": 1.0, \"candidate_ee3275\": 0.0, \"candidate_ced601\": 0.0, \"candidate_8e392d\": 0.0, \"candidate_bb491a\": 0.0, \"candidate_e558d6\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b1491d3223838b52752c7d7a:choice", "group_id": "fresh12:banking77:b1491d3223838b52752c7d7a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I use this account to receive my salary?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5cb0c4\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_b29938\": \"The user's request concerns receiving money.\", \"candidate_91e562\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_602d9a\": \"The user's request concerns verify source of funds.\", \"candidate_51cb01\": \"The user's request concerns exchange charge.\", \"candidate_d84e87\": \"The user's request concerns get physical card.\", \"candidate_0ca1ce\": \"The user's request concerns Refund not showing up.\", \"candidate_63e81f\": \"The user's request concerns disposable card limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5cb0c4\": 0.0, \"candidate_b29938\": 1.0, \"candidate_91e562\": 0.0, \"candidate_602d9a\": 0.0, \"candidate_51cb01\": 0.0, \"candidate_d84e87\": 0.0, \"candidate_0ca1ce\": 0.0, \"candidate_63e81f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6f405603b6c6dba54910f5a5:choice", "group_id": "fresh12:banking77:6f405603b6c6dba54910f5a5", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I get my paycheck through here?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2aba61\": \"The user's request concerns receiving money.\", \"candidate_a6037b\": \"The user's request concerns exchange via app.\", \"candidate_f3843a\": \"The user's request concerns pending card payment.\", \"candidate_cae598\": \"The user's request concerns compromised card.\", \"candidate_602a76\": \"The user's request concerns card delivery estimate.\", \"candidate_e3f84e\": \"The user's request concerns pending transfer.\", \"candidate_a62f44\": \"The user's request concerns why verify identity.\", \"candidate_c52c2f\": \"The user's request concerns get disposable virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2aba61\": 1.0, \"candidate_a6037b\": 0.0, \"candidate_f3843a\": 0.0, \"candidate_cae598\": 0.0, \"candidate_602a76\": 0.0, \"candidate_e3f84e\": 0.0, \"candidate_a62f44\": 0.0, \"candidate_c52c2f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:420e375b172787a652a9a5d5:choice", "group_id": "fresh12:banking77:420e375b172787a652a9a5d5", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can my friend transfer money to me?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_deaf7c\": \"The user's request concerns verify top up.\", \"candidate_765f25\": \"The user's request concerns cash withdrawal charge.\", \"candidate_5008d5\": \"The user's request concerns receiving money.\", \"candidate_2bff94\": \"The user's request concerns getting virtual card.\", \"candidate_5dd414\": \"The user's request concerns top up failed.\", \"candidate_32adf2\": \"The user's request concerns card about to expire.\", \"candidate_89ccfc\": \"The user's request concerns get physical card.\", \"candidate_2a2c8a\": \"The user's request concerns automatic top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_deaf7c\": 0.0, \"candidate_765f25\": 0.0, \"candidate_5008d5\": 1.0, \"candidate_2bff94\": 0.0, \"candidate_5dd414\": 0.0, \"candidate_32adf2\": 0.0, \"candidate_89ccfc\": 0.0, \"candidate_2a2c8a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b6e1e89a67f15560c67ff0e2:choice", "group_id": "fresh12:banking77:b6e1e89a67f15560c67ff0e2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I deposit my paycheck to this account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6ea030\": \"The user's request concerns top up by card charge.\", \"candidate_c885f2\": \"The user's request concerns terminate account.\", \"candidate_dc6440\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_8e3e61\": \"The user's request concerns receiving money.\", \"candidate_3f76f2\": \"The user's request concerns get disposable virtual card.\", \"candidate_612cc0\": \"The user's request concerns atm support.\", \"candidate_8f9002\": \"The user's request concerns declined cash withdrawal.\", \"candidate_fc8361\": \"The user's request concerns card payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6ea030\": 0.0, \"candidate_c885f2\": 0.0, \"candidate_dc6440\": 0.0, \"candidate_8e3e61\": 1.0, \"candidate_3f76f2\": 0.0, \"candidate_612cc0\": 0.0, \"candidate_8f9002\": 0.0, \"candidate_fc8361\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:56a1ce8113f06b05d7b531fe:choice", "group_id": "fresh12:banking77:56a1ce8113f06b05d7b531fe", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can my friends send me money?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_759963\": \"The user's request concerns receiving money.\", \"candidate_96720b\": \"The user's request concerns transfer into account.\", \"candidate_6378f1\": \"The user's request concerns Refund not showing up.\", \"candidate_c26bca\": \"The user's request concerns declined transfer.\", \"candidate_a58ead\": \"The user's request concerns top up failed.\", \"candidate_32c319\": \"The user's request concerns card swallowed.\", \"candidate_38478f\": \"The user's request concerns visa or mastercard.\", \"candidate_b2f618\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_759963\": 1.0, \"candidate_96720b\": 0.0, \"candidate_6378f1\": 0.0, \"candidate_c26bca\": 0.0, \"candidate_a58ead\": 0.0, \"candidate_32c319\": 0.0, \"candidate_38478f\": 0.0, \"candidate_b2f618\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:79d3162582569474205482cb:choice", "group_id": "fresh12:banking77:79d3162582569474205482cb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I transferred my funds, why did it not go through?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1f9952\": \"The user's request concerns top up by card charge.\", \"candidate_229440\": \"The user's request concerns edit personal details.\", \"candidate_546d78\": \"The user's request concerns top up failed.\", \"candidate_a9f504\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_ef5659\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_f92537\": \"The user's request concerns receiving money.\", \"candidate_11dc40\": \"The user's request concerns transfer into account.\", \"candidate_dcdc26\": \"The user's request concerns failed transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1f9952\": 0.0, \"candidate_229440\": 0.0, \"candidate_546d78\": 0.0, \"candidate_a9f504\": 0.0, \"candidate_ef5659\": 0.0, \"candidate_f92537\": 0.0, \"candidate_11dc40\": 0.0, \"candidate_dcdc26\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:503fd9a7ff6f346d3cd19324:choice", "group_id": "fresh12:banking77:503fd9a7ff6f346d3cd19324", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is the system down? I've been trying to do a transfer and have failed?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_20d1cc\": \"The user's request concerns transaction charged twice.\", \"candidate_38b454\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_0c5497\": \"The user's request concerns declined card payment.\", \"candidate_d3a262\": \"The user's request concerns failed transfer.\", \"candidate_759273\": \"The user's request concerns request refund.\", \"candidate_1bb830\": \"The user's request concerns card acceptance.\", \"candidate_985aa2\": \"The user's request concerns exchange via app.\", \"candidate_f4836f\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_20d1cc\": 0.0, \"candidate_38b454\": 0.0, \"candidate_0c5497\": 0.0, \"candidate_d3a262\": 1.0, \"candidate_759273\": 0.0, \"candidate_1bb830\": 0.0, \"candidate_985aa2\": 0.0, \"candidate_f4836f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1c8c2cc9cd5cc785dc6ef9ef:choice", "group_id": "fresh12:banking77:1c8c2cc9cd5cc785dc6ef9ef", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"how come my transfer didn't work\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e08c72\": \"The user's request concerns transaction charged twice.\", \"candidate_9671f3\": \"The user's request concerns supported cards and currencies.\", \"candidate_d26201\": \"The user's request concerns disposable card limits.\", \"candidate_8dd3e5\": \"The user's request concerns exchange rate.\", \"candidate_5cd63c\": \"The user's request concerns failed transfer.\", \"candidate_86f1f5\": \"The user's request concerns verify top up.\", \"candidate_dcf20e\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_6957cf\": \"The user's request concerns card arrival.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e08c72\": 0.0, \"candidate_9671f3\": 0.0, \"candidate_d26201\": 0.0, \"candidate_8dd3e5\": 0.0, \"candidate_5cd63c\": 1.0, \"candidate_86f1f5\": 0.0, \"candidate_dcf20e\": 0.0, \"candidate_6957cf\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2e482e930f8a2088fd99ebc4:choice", "group_id": "fresh12:banking77:2e482e930f8a2088fd99ebc4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My transfer didn't go through.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_396c7f\": \"The user's request concerns age limit.\", \"candidate_5820e2\": \"The user's request concerns virtual card not working.\", \"candidate_bdb810\": \"The user's request concerns transfer fee charged.\", \"candidate_812037\": \"The user's request concerns visa or mastercard.\", \"candidate_7e7dfb\": \"The user's request concerns get physical card.\", \"candidate_28aa0e\": \"The user's request concerns failed transfer.\", \"candidate_b38c40\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_8d2cc6\": \"The user's request concerns transfer timing.\"}}}}", "targets_json": "{\"decision\": {\"candidate_396c7f\": 0.0, \"candidate_5820e2\": 0.0, \"candidate_bdb810\": 0.0, \"candidate_812037\": 0.0, \"candidate_7e7dfb\": 0.0, \"candidate_28aa0e\": 1.0, \"candidate_b38c40\": 0.0, \"candidate_8d2cc6\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e7ccaaed6ce6991ce7427bc9:choice", "group_id": "fresh12:banking77:e7ccaaed6ce6991ce7427bc9", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The transfer keeps failing , I tried to transfer some money to friends this morning but it keeps getting rejected for some reason, Would you please check the issue ?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_49a04b\": \"The user's request concerns why verify identity.\", \"candidate_c9b912\": \"The user's request concerns verify top up.\", \"candidate_00cea3\": \"The user's request concerns failed transfer.\", \"candidate_a16489\": \"The user's request concerns disposable card limits.\", \"candidate_400ae3\": \"The user's request concerns virtual card not working.\", \"candidate_927a49\": \"The user's request concerns lost or stolen card.\", \"candidate_2e8690\": \"The user's request concerns unable to verify identity.\", \"candidate_b7cbcf\": \"The user's request concerns visa or mastercard.\"}}}}", "targets_json": "{\"decision\": {\"candidate_49a04b\": 0.0, \"candidate_c9b912\": 0.0, \"candidate_00cea3\": 1.0, \"candidate_a16489\": 0.0, \"candidate_400ae3\": 0.0, \"candidate_927a49\": 0.0, \"candidate_2e8690\": 0.0, \"candidate_b7cbcf\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3a84c9233f942de530350174:choice", "group_id": "fresh12:banking77:3a84c9233f942de530350174", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Tell me why my transfer failed!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1ee399\": \"The user's request concerns failed transfer.\", \"candidate_fa6cf2\": \"The user's request concerns automatic top up.\", \"candidate_c54f1e\": \"The user's request concerns change pin.\", \"candidate_b116c0\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_039857\": \"The user's request concerns declined transfer.\", \"candidate_8197a5\": \"The user's request concerns unable to verify identity.\", \"candidate_23bf19\": \"The user's request concerns card acceptance.\", \"candidate_daee05\": \"The user's request concerns transaction charged twice.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1ee399\": 1.0, \"candidate_fa6cf2\": 0.0, \"candidate_c54f1e\": 0.0, \"candidate_b116c0\": 0.0, \"candidate_039857\": 0.0, \"candidate_8197a5\": 0.0, \"candidate_23bf19\": 0.0, \"candidate_daee05\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0fd94c1f13acdd22d1545032:choice", "group_id": "fresh12:banking77:0fd94c1f13acdd22d1545032", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need to bank account transfer from another account into this one, what is the easiest way that can be done, and what information do i need?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9fd0ca\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_66226d\": \"The user's request concerns getting virtual card.\", \"candidate_a87211\": \"The user's request concerns transfer not received by recipient.\", \"candidate_7d8507\": \"The user's request concerns transfer into account.\", \"candidate_fbafac\": \"The user's request concerns visa or mastercard.\", \"candidate_9e472c\": \"The user's request concerns get physical card.\", \"candidate_fc44d2\": \"The user's request concerns pending top up.\", \"candidate_7625cc\": \"The user's request concerns request refund.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9fd0ca\": 0.0, \"candidate_66226d\": 0.0, \"candidate_a87211\": 0.0, \"candidate_7d8507\": 1.0, \"candidate_fbafac\": 0.0, \"candidate_9e472c\": 0.0, \"candidate_fc44d2\": 0.0, \"candidate_7625cc\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:cdddecdd83b327e7c5a18720:choice", "group_id": "fresh12:banking77:cdddecdd83b327e7c5a18720", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What do I need to do to transfer money into my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_aa9aae\": \"The user's request concerns declined card payment.\", \"candidate_23cd5b\": \"The user's request concerns top up by card charge.\", \"candidate_82d0b0\": \"The user's request concerns card payment not recognised.\", \"candidate_a800de\": \"The user's request concerns card delivery estimate.\", \"candidate_c27dde\": \"The user's request concerns change pin.\", \"candidate_a6296e\": \"The user's request concerns transfer into account.\", \"candidate_f379ee\": \"The user's request concerns request refund.\", \"candidate_b6b6f7\": \"The user's request concerns cash withdrawal charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_aa9aae\": 0.0, \"candidate_23cd5b\": 0.0, \"candidate_82d0b0\": 0.0, \"candidate_a800de\": 0.0, \"candidate_c27dde\": 0.0, \"candidate_a6296e\": 1.0, \"candidate_f379ee\": 0.0, \"candidate_b6b6f7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b023896e265afdd0a0624793:choice", "group_id": "fresh12:banking77:b023896e265afdd0a0624793", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I would like to transfer money from my other bank account into this one.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f3f145\": \"The user's request concerns card payment fee charged.\", \"candidate_196db2\": \"The user's request concerns transfer into account.\", \"candidate_661d38\": \"The user's request concerns card linking.\", \"candidate_2eb313\": \"The user's request concerns pending card payment.\", \"candidate_1e9999\": \"The user's request concerns pin blocked.\", \"candidate_94f291\": \"The user's request concerns declined card payment.\", \"candidate_c1bc34\": \"The user's request concerns terminate account.\", \"candidate_e301d7\": \"The user's request concerns Refund not showing up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f3f145\": 0.0, \"candidate_196db2\": 1.0, \"candidate_661d38\": 0.0, \"candidate_2eb313\": 0.0, \"candidate_1e9999\": 0.0, \"candidate_94f291\": 0.0, \"candidate_c1bc34\": 0.0, \"candidate_e301d7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e00df01093c5439b57cdb589:choice", "group_id": "fresh12:banking77:e00df01093c5439b57cdb589", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How to I transfer my funds into my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cbed1c\": \"The user's request concerns terminate account.\", \"candidate_a8ee94\": \"The user's request concerns verify source of funds.\", \"candidate_926166\": \"The user's request concerns transfer into account.\", \"candidate_fc716f\": \"The user's request concerns edit personal details.\", \"candidate_dfc9fa\": \"The user's request concerns verify my identity.\", \"candidate_af6722\": \"The user's request concerns declined card payment.\", \"candidate_d424a5\": \"The user's request concerns wrong amount of cash received.\", \"candidate_19fb25\": \"The user's request concerns balance not updated after cheque or cash deposit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cbed1c\": 0.0, \"candidate_a8ee94\": 0.0, \"candidate_926166\": 1.0, \"candidate_fc716f\": 0.0, \"candidate_dfc9fa\": 0.0, \"candidate_af6722\": 0.0, \"candidate_d424a5\": 0.0, \"candidate_19fb25\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ad85dbdd4a29da9472a6c11c:choice", "group_id": "fresh12:banking77:ad85dbdd4a29da9472a6c11c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I don't know what to do. Should I transfer funds. My account is out of money.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b4fb9d\": \"The user's request concerns visa or mastercard.\", \"candidate_2977ae\": \"The user's request concerns pending transfer.\", \"candidate_e68bbd\": \"The user's request concerns pending card payment.\", \"candidate_f88fe3\": \"The user's request concerns age limit.\", \"candidate_e55b46\": \"The user's request concerns transfer fee charged.\", \"candidate_9dd36c\": \"The user's request concerns transfer into account.\", \"candidate_e59da6\": \"The user's request concerns Refund not showing up.\", \"candidate_990253\": \"The user's request concerns lost or stolen card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b4fb9d\": 0.0, \"candidate_2977ae\": 0.0, \"candidate_e68bbd\": 0.0, \"candidate_f88fe3\": 0.0, \"candidate_e55b46\": 0.0, \"candidate_9dd36c\": 1.0, \"candidate_e59da6\": 0.0, \"candidate_990253\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ee7cc203653ec70b5e144617:choice", "group_id": "fresh12:banking77:ee7cc203653ec70b5e144617", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"When I want to transfer money to my account, how can I do that?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_833d91\": \"The user's request concerns exchange rate.\", \"candidate_56cbd8\": \"The user's request concerns card about to expire.\", \"candidate_b5ac7b\": \"The user's request concerns edit personal details.\", \"candidate_4c1f91\": \"The user's request concerns card arrival.\", \"candidate_f307e9\": \"The user's request concerns lost or stolen card.\", \"candidate_bd9e5f\": \"The user's request concerns transfer into account.\", \"candidate_34c326\": \"The user's request concerns pending card payment.\", \"candidate_e65de6\": \"The user's request concerns reverted card payment?.\"}}}}", "targets_json": "{\"decision\": {\"candidate_833d91\": 0.0, \"candidate_56cbd8\": 0.0, \"candidate_b5ac7b\": 0.0, \"candidate_4c1f91\": 0.0, \"candidate_f307e9\": 0.0, \"candidate_bd9e5f\": 1.0, \"candidate_34c326\": 0.0, \"candidate_e65de6\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c5fd801ab8d03ea0860b4db0:choice", "group_id": "fresh12:banking77:c5fd801ab8d03ea0860b4db0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I can't find the top-up verification code.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_65dbda\": \"The user's request concerns verify top up.\", \"candidate_9840d4\": \"The user's request concerns card arrival.\", \"candidate_2fdd4c\": \"The user's request concerns declined transfer.\", \"candidate_254d5e\": \"The user's request concerns pending cash withdrawal.\", \"candidate_ff6a74\": \"The user's request concerns country support.\", \"candidate_1ce719\": \"The user's request concerns why verify identity.\", \"candidate_8c22eb\": \"The user's request concerns card about to expire.\", \"candidate_d076e8\": \"The user's request concerns cash withdrawal not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_65dbda\": 1.0, \"candidate_9840d4\": 0.0, \"candidate_2fdd4c\": 0.0, \"candidate_254d5e\": 0.0, \"candidate_ff6a74\": 0.0, \"candidate_1ce719\": 0.0, \"candidate_8c22eb\": 0.0, \"candidate_d076e8\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2aeff45b7b13f4cf5219c3db:choice", "group_id": "fresh12:banking77:2aeff45b7b13f4cf5219c3db", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why am i to verify the top-ups?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_076699\": \"The user's request concerns failed transfer.\", \"candidate_8cb926\": \"The user's request concerns verify my identity.\", \"candidate_e14dde\": \"The user's request concerns verify top up.\", \"candidate_f9fc1c\": \"The user's request concerns apple pay or google pay.\", \"candidate_dd21c1\": \"The user's request concerns fiat currency support.\", \"candidate_d39e8e\": \"The user's request concerns top up reverted.\", \"candidate_95dbf1\": \"The user's request concerns getting virtual card.\", \"candidate_698689\": \"The user's request concerns virtual card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_076699\": 0.0, \"candidate_8cb926\": 0.0, \"candidate_e14dde\": 1.0, \"candidate_f9fc1c\": 0.0, \"candidate_dd21c1\": 0.0, \"candidate_d39e8e\": 0.0, \"candidate_95dbf1\": 0.0, \"candidate_698689\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:445f368cfa81aef4ef5c1a05:choice", "group_id": "fresh12:banking77:445f368cfa81aef4ef5c1a05", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why does top-up need verifying?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d3605b\": \"The user's request concerns fiat currency support.\", \"candidate_ae9e66\": \"The user's request concerns reverted card payment?.\", \"candidate_6bf839\": \"The user's request concerns edit personal details.\", \"candidate_72a87d\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_d8b8eb\": \"The user's request concerns compromised card.\", \"candidate_287999\": \"The user's request concerns card swallowed.\", \"candidate_97506d\": \"The user's request concerns verify top up.\", \"candidate_f328c8\": \"The user's request concerns top up by bank transfer charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d3605b\": 0.0, \"candidate_ae9e66\": 0.0, \"candidate_6bf839\": 0.0, \"candidate_72a87d\": 0.0, \"candidate_d8b8eb\": 0.0, \"candidate_287999\": 0.0, \"candidate_97506d\": 1.0, \"candidate_f328c8\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:eb05000e400ca94c2cc13ceb:choice", "group_id": "fresh12:banking77:eb05000e400ca94c2cc13ceb", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why do I have to verify the top-up that I started?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2ccd35\": \"The user's request concerns country support.\", \"candidate_658f8b\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_191a3d\": \"The user's request concerns card acceptance.\", \"candidate_326220\": \"The user's request concerns top up reverted.\", \"candidate_5878d7\": \"The user's request concerns verify top up.\", \"candidate_06c15b\": \"The user's request concerns top up failed.\", \"candidate_517ee0\": \"The user's request concerns pending card payment.\", \"candidate_e2e08c\": \"The user's request concerns direct debit payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2ccd35\": 0.0, \"candidate_658f8b\": 0.0, \"candidate_191a3d\": 0.0, \"candidate_326220\": 0.0, \"candidate_5878d7\": 1.0, \"candidate_06c15b\": 0.0, \"candidate_517ee0\": 0.0, \"candidate_e2e08c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9b0812af42e2d1ff705082b0:choice", "group_id": "fresh12:banking77:9b0812af42e2d1ff705082b0", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a verification code for top-ups?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1a2ed4\": \"The user's request concerns pending transfer.\", \"candidate_fea635\": \"The user's request concerns transfer timing.\", \"candidate_976463\": \"The user's request concerns change pin.\", \"candidate_c9ad97\": \"The user's request concerns cancel transfer.\", \"candidate_8621a2\": \"The user's request concerns activate my card.\", \"candidate_eae937\": \"The user's request concerns extra charge on statement.\", \"candidate_c72f57\": \"The user's request concerns lost or stolen phone.\", \"candidate_820413\": \"The user's request concerns verify top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1a2ed4\": 0.0, \"candidate_fea635\": 0.0, \"candidate_976463\": 0.0, \"candidate_c9ad97\": 0.0, \"candidate_8621a2\": 0.0, \"candidate_eae937\": 0.0, \"candidate_c72f57\": 0.0, \"candidate_820413\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:16be8589b243e777d102a883:choice", "group_id": "fresh12:banking77:16be8589b243e777d102a883", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why do you need code verification for my top up?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5ce821\": \"The user's request concerns card acceptance.\", \"candidate_9ab2e0\": \"The user's request concerns wrong amount of cash received.\", \"candidate_950347\": \"The user's request concerns disposable card limits.\", \"candidate_03ae53\": \"The user's request concerns get physical card.\", \"candidate_b2a4b2\": \"The user's request concerns verify top up.\", \"candidate_b346e9\": \"The user's request concerns top up failed.\", \"candidate_72b8e2\": \"The user's request concerns passcode forgotten.\", \"candidate_9fe697\": \"The user's request concerns Refund not showing up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5ce821\": 0.0, \"candidate_9ab2e0\": 0.0, \"candidate_950347\": 0.0, \"candidate_03ae53\": 0.0, \"candidate_b2a4b2\": 1.0, \"candidate_b346e9\": 0.0, \"candidate_72b8e2\": 0.0, \"candidate_9fe697\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3d2a99020722c633c2365d6b:choice", "group_id": "fresh12:banking77:3d2a99020722c633c2365d6b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a fee for sending out more than one card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cd593f\": \"The user's request concerns automatic top up.\", \"candidate_28499a\": \"The user's request concerns cancel transfer.\", \"candidate_d810dc\": \"The user's request concerns card payment fee charged.\", \"candidate_31f163\": \"The user's request concerns card swallowed.\", \"candidate_ce3125\": \"The user's request concerns pending transfer.\", \"candidate_4656bb\": \"The user's request concerns extra charge on statement.\", \"candidate_519fb0\": \"The user's request concerns getting spare card.\", \"candidate_f1c568\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cd593f\": 0.0, \"candidate_28499a\": 0.0, \"candidate_d810dc\": 0.0, \"candidate_31f163\": 0.0, \"candidate_ce3125\": 0.0, \"candidate_4656bb\": 0.0, \"candidate_519fb0\": 1.0, \"candidate_f1c568\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:544bd6813b599d8d804cd463:choice", "group_id": "fresh12:banking77:544bd6813b599d8d804cd463", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"If you send me more cards, are there any charges?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d3caf2\": \"The user's request concerns reverted card payment?.\", \"candidate_d27f9a\": \"The user's request concerns lost or stolen phone.\", \"candidate_9ec060\": \"The user's request concerns pin blocked.\", \"candidate_f3e0a2\": \"The user's request concerns atm support.\", \"candidate_d5a78c\": \"The user's request concerns age limit.\", \"candidate_7226c0\": \"The user's request concerns getting spare card.\", \"candidate_e66a77\": \"The user's request concerns transfer fee charged.\", \"candidate_bf7caf\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d3caf2\": 0.0, \"candidate_d27f9a\": 0.0, \"candidate_9ec060\": 0.0, \"candidate_f3e0a2\": 0.0, \"candidate_d5a78c\": 0.0, \"candidate_7226c0\": 1.0, \"candidate_e66a77\": 0.0, \"candidate_bf7caf\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b9fec1aed5ffa81f2e56dc60:choice", "group_id": "fresh12:banking77:b9fec1aed5ffa81f2e56dc60", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What is the procedure for getting another card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_dd838f\": \"The user's request concerns exchange charge.\", \"candidate_aca802\": \"The user's request concerns card not working.\", \"candidate_2991a9\": \"The user's request concerns getting spare card.\", \"candidate_08d607\": \"The user's request concerns exchange via app.\", \"candidate_ef1692\": \"The user's request concerns declined card payment.\", \"candidate_60ec85\": \"The user's request concerns edit personal details.\", \"candidate_3df0a9\": \"The user's request concerns top up reverted.\", \"candidate_d59d86\": \"The user's request concerns reverted card payment?.\"}}}}", "targets_json": "{\"decision\": {\"candidate_dd838f\": 0.0, \"candidate_aca802\": 0.0, \"candidate_2991a9\": 1.0, \"candidate_08d607\": 0.0, \"candidate_ef1692\": 0.0, \"candidate_60ec85\": 0.0, \"candidate_3df0a9\": 0.0, \"candidate_d59d86\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a84bbead4d0133e5ac1b0733:choice", "group_id": "fresh12:banking77:a84bbead4d0133e5ac1b0733", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Would I be able to get another card for my account, so I could give one to my daughter?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_efe028\": \"The user's request concerns supported cards and currencies.\", \"candidate_1bb41f\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_6c843f\": \"The user's request concerns getting spare card.\", \"candidate_50cc6f\": \"The user's request concerns Refund not showing up.\", \"candidate_12368a\": \"The user's request concerns card delivery estimate.\", \"candidate_b1d71d\": \"The user's request concerns get disposable virtual card.\", \"candidate_7ec6ce\": \"The user's request concerns topping up by card.\", \"candidate_9f94a9\": \"The user's request concerns transaction charged twice.\"}}}}", "targets_json": "{\"decision\": {\"candidate_efe028\": 0.0, \"candidate_1bb41f\": 0.0, \"candidate_6c843f\": 1.0, \"candidate_50cc6f\": 0.0, \"candidate_12368a\": 0.0, \"candidate_b1d71d\": 0.0, \"candidate_7ec6ce\": 0.0, \"candidate_9f94a9\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e3a90c562db4ed682a86a962:choice", "group_id": "fresh12:banking77:e3a90c562db4ed682a86a962", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How can I get additional cards?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b407d5\": \"The user's request concerns edit personal details.\", \"candidate_457947\": \"The user's request concerns top up failed.\", \"candidate_b34b62\": \"The user's request concerns exchange rate.\", \"candidate_8bafc9\": \"The user's request concerns receiving money.\", \"candidate_654663\": \"The user's request concerns order physical card.\", \"candidate_4225f0\": \"The user's request concerns getting spare card.\", \"candidate_13472b\": \"The user's request concerns pending top up.\", \"candidate_f06c31\": \"The user's request concerns card payment fee charged.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b407d5\": 0.0, \"candidate_457947\": 0.0, \"candidate_b34b62\": 0.0, \"candidate_8bafc9\": 0.0, \"candidate_654663\": 0.0, \"candidate_4225f0\": 1.0, \"candidate_13472b\": 0.0, \"candidate_f06c31\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c29b8e26a9e2ce57078d0c1f:choice", "group_id": "fresh12:banking77:c29b8e26a9e2ce57078d0c1f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I get a spare card for someone else to use?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_308b00\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_ab6c59\": \"The user's request concerns cash withdrawal charge.\", \"candidate_78e9f7\": \"The user's request concerns virtual card not working.\", \"candidate_4386ed\": \"The user's request concerns failed transfer.\", \"candidate_cc3ce0\": \"The user's request concerns getting spare card.\", \"candidate_96965a\": \"The user's request concerns lost or stolen phone.\", \"candidate_9cb13d\": \"The user's request concerns exchange charge.\", \"candidate_54dd5e\": \"The user's request concerns top up by card charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_308b00\": 0.0, \"candidate_ab6c59\": 0.0, \"candidate_78e9f7\": 0.0, \"candidate_4386ed\": 0.0, \"candidate_cc3ce0\": 1.0, \"candidate_96965a\": 0.0, \"candidate_9cb13d\": 0.0, \"candidate_54dd5e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7590cedc4e64f0190da82e83:choice", "group_id": "fresh12:banking77:7590cedc4e64f0190da82e83", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I want to do a cash top-up\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_eb80f4\": \"The user's request concerns top up reverted.\", \"candidate_322daa\": \"The user's request concerns topping up by card.\", \"candidate_dd8e12\": \"The user's request concerns top up by cash or cheque.\", \"candidate_9a590f\": \"The user's request concerns declined transfer.\", \"candidate_2d1da8\": \"The user's request concerns unable to verify identity.\", \"candidate_3cd705\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_75f7e6\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_98c8ba\": \"The user's request concerns pending transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_eb80f4\": 0.0, \"candidate_322daa\": 0.0, \"candidate_dd8e12\": 1.0, \"candidate_9a590f\": 0.0, \"candidate_2d1da8\": 0.0, \"candidate_3cd705\": 0.0, \"candidate_75f7e6\": 0.0, \"candidate_98c8ba\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:130239bad1c68fa8804c6654:choice", "group_id": "fresh12:banking77:130239bad1c68fa8804c6654", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"how come i can't find anywhere to load using cash\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_fed225\": \"The user's request concerns get disposable virtual card.\", \"candidate_eafd96\": \"The user's request concerns exchange charge.\", \"candidate_51248f\": \"The user's request concerns contactless not working.\", \"candidate_0b1ccc\": \"The user's request concerns automatic top up.\", \"candidate_ebce9f\": \"The user's request concerns request refund.\", \"candidate_edc16a\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_462c00\": \"The user's request concerns extra charge on statement.\", \"candidate_b19a19\": \"The user's request concerns top up by cash or cheque.\"}}}}", "targets_json": "{\"decision\": {\"candidate_fed225\": 0.0, \"candidate_eafd96\": 0.0, \"candidate_51248f\": 0.0, \"candidate_0b1ccc\": 0.0, \"candidate_ebce9f\": 0.0, \"candidate_edc16a\": 0.0, \"candidate_462c00\": 0.0, \"candidate_b19a19\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f7fc07c4471f7929e5520fb2:choice", "group_id": "fresh12:banking77:f7fc07c4471f7929e5520fb2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is it possible to top up with cash? If so how do I do it?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e0422b\": \"The user's request concerns exchange via app.\", \"candidate_42d6d2\": \"The user's request concerns card swallowed.\", \"candidate_bc0def\": \"The user's request concerns getting spare card.\", \"candidate_3eec08\": \"The user's request concerns top up by cash or cheque.\", \"candidate_76fbf7\": \"The user's request concerns automatic top up.\", \"candidate_9d3059\": \"The user's request concerns supported cards and currencies.\", \"candidate_e7d0ec\": \"The user's request concerns get disposable virtual card.\", \"candidate_8dc73a\": \"The user's request concerns declined card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e0422b\": 0.0, \"candidate_42d6d2\": 0.0, \"candidate_bc0def\": 0.0, \"candidate_3eec08\": 1.0, \"candidate_76fbf7\": 0.0, \"candidate_9d3059\": 0.0, \"candidate_e7d0ec\": 0.0, \"candidate_8dc73a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:23dd9ec38746b6b5aee8f9af:choice", "group_id": "fresh12:banking77:23dd9ec38746b6b5aee8f9af", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I top-up using cash?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3927e9\": \"The user's request concerns card linking.\", \"candidate_5bdb2a\": \"The user's request concerns verify my identity.\", \"candidate_400946\": \"The user's request concerns transaction charged twice.\", \"candidate_12c1c3\": \"The user's request concerns extra charge on statement.\", \"candidate_31f941\": \"The user's request concerns top up by cash or cheque.\", \"candidate_8b505d\": \"The user's request concerns country support.\", \"candidate_faeb8b\": \"The user's request concerns verify top up.\", \"candidate_3b39be\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3927e9\": 0.0, \"candidate_5bdb2a\": 0.0, \"candidate_400946\": 0.0, \"candidate_12c1c3\": 0.0, \"candidate_31f941\": 1.0, \"candidate_8b505d\": 0.0, \"candidate_faeb8b\": 0.0, \"candidate_3b39be\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1ad6b598feefcb60fb7183f3:choice", "group_id": "fresh12:banking77:1ad6b598feefcb60fb7183f3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I send a cheque to do my top ups?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9c7438\": \"The user's request concerns top up by cash or cheque.\", \"candidate_b7c922\": \"The user's request concerns disposable card limits.\", \"candidate_7add43\": \"The user's request concerns exchange rate.\", \"candidate_e81002\": \"The user's request concerns verify top up.\", \"candidate_b39405\": \"The user's request concerns card about to expire.\", \"candidate_2b895a\": \"The user's request concerns virtual card not working.\", \"candidate_4fffbd\": \"The user's request concerns contactless not working.\", \"candidate_ea6b04\": \"The user's request concerns top up failed.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9c7438\": 1.0, \"candidate_b7c922\": 0.0, \"candidate_7add43\": 0.0, \"candidate_e81002\": 0.0, \"candidate_b39405\": 0.0, \"candidate_2b895a\": 0.0, \"candidate_4fffbd\": 0.0, \"candidate_ea6b04\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8551e095433c4bc57d679f92:choice", "group_id": "fresh12:banking77:8551e095433c4bc57d679f92", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why can't I top up with a cheque?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_989209\": \"The user's request concerns top up by cash or cheque.\", \"candidate_bc120d\": \"The user's request concerns passcode forgotten.\", \"candidate_9a0956\": \"The user's request concerns apple pay or google pay.\", \"candidate_22dca0\": \"The user's request concerns contactless not working.\", \"candidate_acb16b\": \"The user's request concerns declined card payment.\", \"candidate_dad906\": \"The user's request concerns topping up by card.\", \"candidate_bd5f6b\": \"The user's request concerns activate my card.\", \"candidate_bc9f38\": \"The user's request concerns wrong amount of cash received.\"}}}}", "targets_json": "{\"decision\": {\"candidate_989209\": 1.0, \"candidate_bc120d\": 0.0, \"candidate_9a0956\": 0.0, \"candidate_22dca0\": 0.0, \"candidate_acb16b\": 0.0, \"candidate_dad906\": 0.0, \"candidate_bd5f6b\": 0.0, \"candidate_bc9f38\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a0e07569867b76407f674c9a:choice", "group_id": "fresh12:banking77:a0e07569867b76407f674c9a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Do you charge for physical cards\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_8d6a66\": \"The user's request concerns edit personal details.\", \"candidate_cc19b3\": \"The user's request concerns topping up by card.\", \"candidate_fcf366\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_30c0a6\": \"The user's request concerns card linking.\", \"candidate_82a72e\": \"The user's request concerns transfer not received by recipient.\", \"candidate_4b5f48\": \"The user's request concerns declined transfer.\", \"candidate_ab1a63\": \"The user's request concerns pending top up.\", \"candidate_71161e\": \"The user's request concerns order physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_8d6a66\": 0.0, \"candidate_cc19b3\": 0.0, \"candidate_fcf366\": 0.0, \"candidate_30c0a6\": 0.0, \"candidate_82a72e\": 0.0, \"candidate_4b5f48\": 0.0, \"candidate_ab1a63\": 0.0, \"candidate_71161e\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3a2ca622c34437c5ab014a72:choice", "group_id": "fresh12:banking77:3a2ca622c34437c5ab014a72", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Send me a physical card\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2d84de\": \"The user's request concerns lost or stolen card.\", \"candidate_21e545\": \"The user's request concerns transfer not received by recipient.\", \"candidate_4c0ed9\": \"The user's request concerns transfer into account.\", \"candidate_32dc36\": \"The user's request concerns declined cash withdrawal.\", \"candidate_bae3ca\": \"The user's request concerns order physical card.\", \"candidate_9a7581\": \"The user's request concerns wrong amount of cash received.\", \"candidate_e4f156\": \"The user's request concerns lost or stolen phone.\", \"candidate_2a6759\": \"The user's request concerns top up limits.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2d84de\": 0.0, \"candidate_21e545\": 0.0, \"candidate_4c0ed9\": 0.0, \"candidate_32dc36\": 0.0, \"candidate_bae3ca\": 1.0, \"candidate_9a7581\": 0.0, \"candidate_e4f156\": 0.0, \"candidate_2a6759\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:88f9423417f6220293bbda6c:choice", "group_id": "fresh12:banking77:88f9423417f6220293bbda6c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What are the fees for a physical card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f77196\": \"The user's request concerns pin blocked.\", \"candidate_2cdf97\": \"The user's request concerns verify top up.\", \"candidate_a44557\": \"The user's request concerns top up limits.\", \"candidate_5fa9dd\": \"The user's request concerns card about to expire.\", \"candidate_dd49b7\": \"The user's request concerns order physical card.\", \"candidate_041514\": \"The user's request concerns get disposable virtual card.\", \"candidate_deca9b\": \"The user's request concerns activate my card.\", \"candidate_2d7c90\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f77196\": 0.0, \"candidate_2cdf97\": 0.0, \"candidate_a44557\": 0.0, \"candidate_5fa9dd\": 0.0, \"candidate_dd49b7\": 1.0, \"candidate_041514\": 0.0, \"candidate_deca9b\": 0.0, \"candidate_2d7c90\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:773254748eb2c9a5ea099ee5:choice", "group_id": "fresh12:banking77:773254748eb2c9a5ea099ee5", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I want a card. What is the procedure?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d963bc\": \"The user's request concerns get disposable virtual card.\", \"candidate_3a51cf\": \"The user's request concerns visa or mastercard.\", \"candidate_22c543\": \"The user's request concerns order physical card.\", \"candidate_5e1dca\": \"The user's request concerns exchange rate.\", \"candidate_89a7d4\": \"The user's request concerns receiving money.\", \"candidate_a1f4c1\": \"The user's request concerns verify source of funds.\", \"candidate_3f8f2e\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_c02383\": \"The user's request concerns card delivery estimate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d963bc\": 0.0, \"candidate_3a51cf\": 0.0, \"candidate_22c543\": 1.0, \"candidate_5e1dca\": 0.0, \"candidate_89a7d4\": 0.0, \"candidate_a1f4c1\": 0.0, \"candidate_3f8f2e\": 0.0, \"candidate_c02383\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:aa2a3d7dc6adf88e550ba8cc:choice", "group_id": "fresh12:banking77:aa2a3d7dc6adf88e550ba8cc", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How much does it cost in fees to use your card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2e4c6d\": \"The user's request concerns automatic top up.\", \"candidate_01340c\": \"The user's request concerns edit personal details.\", \"candidate_dfa704\": \"The user's request concerns apple pay or google pay.\", \"candidate_8b0813\": \"The user's request concerns transfer timing.\", \"candidate_86a397\": \"The user's request concerns transfer into account.\", \"candidate_485e68\": \"The user's request concerns order physical card.\", \"candidate_479bec\": \"The user's request concerns disposable card limits.\", \"candidate_7daf03\": \"The user's request concerns get disposable virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2e4c6d\": 0.0, \"candidate_01340c\": 0.0, \"candidate_dfa704\": 0.0, \"candidate_8b0813\": 0.0, \"candidate_86a397\": 0.0, \"candidate_485e68\": 1.0, \"candidate_479bec\": 0.0, \"candidate_7daf03\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3698165a41b0edc46df21513:choice", "group_id": "fresh12:banking77:3698165a41b0edc46df21513", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a charge for physical cards\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f975c1\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_bb69ce\": \"The user's request concerns order physical card.\", \"candidate_d626d0\": \"The user's request concerns reverted card payment?.\", \"candidate_997a93\": \"The user's request concerns getting virtual card.\", \"candidate_73663a\": \"The user's request concerns verify top up.\", \"candidate_c1405e\": \"The user's request concerns card linking.\", \"candidate_5aed04\": \"The user's request concerns visa or mastercard.\", \"candidate_d1f04b\": \"The user's request concerns balance not updated after bank transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f975c1\": 0.0, \"candidate_bb69ce\": 1.0, \"candidate_d626d0\": 0.0, \"candidate_997a93\": 0.0, \"candidate_73663a\": 0.0, \"candidate_c1405e\": 0.0, \"candidate_5aed04\": 0.0, \"candidate_d1f04b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2d98a17b3a57b0d9be427b50:choice", "group_id": "fresh12:banking77:2d98a17b3a57b0d9be427b50", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My virtual card won't work.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_b2ed21\": \"The user's request concerns wrong amount of cash received.\", \"candidate_8de90e\": \"The user's request concerns verify top up.\", \"candidate_a4860b\": \"The user's request concerns virtual card not working.\", \"candidate_7b56af\": \"The user's request concerns compromised card.\", \"candidate_b6f824\": \"The user's request concerns get physical card.\", \"candidate_062971\": \"The user's request concerns atm support.\", \"candidate_78e25a\": \"The user's request concerns visa or mastercard.\", \"candidate_ac8b32\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_b2ed21\": 0.0, \"candidate_8de90e\": 0.0, \"candidate_a4860b\": 1.0, \"candidate_7b56af\": 0.0, \"candidate_b6f824\": 0.0, \"candidate_062971\": 0.0, \"candidate_78e25a\": 0.0, \"candidate_ac8b32\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c1f5aea854c05ed731cca33a:choice", "group_id": "fresh12:banking77:c1f5aea854c05ed731cca33a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Are there restrictions for my disposable card since it does not seem to be working?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_50dc27\": \"The user's request concerns card payment not recognised.\", \"candidate_9d94f1\": \"The user's request concerns activate my card.\", \"candidate_540ec4\": \"The user's request concerns transfer into account.\", \"candidate_f3c703\": \"The user's request concerns pin blocked.\", \"candidate_68ae1e\": \"The user's request concerns virtual card not working.\", \"candidate_11f14a\": \"The user's request concerns transfer not received by recipient.\", \"candidate_a4661a\": \"The user's request concerns wrong amount of cash received.\", \"candidate_fea206\": \"The user's request concerns verify top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_50dc27\": 0.0, \"candidate_9d94f1\": 0.0, \"candidate_540ec4\": 0.0, \"candidate_f3c703\": 0.0, \"candidate_68ae1e\": 1.0, \"candidate_11f14a\": 0.0, \"candidate_a4661a\": 0.0, \"candidate_fea206\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6e64133890153b8b33e7b61a:choice", "group_id": "fresh12:banking77:6e64133890153b8b33e7b61a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My disposable virtual card doesn't seem to work\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0fa176\": \"The user's request concerns pending card payment.\", \"candidate_6754ac\": \"The user's request concerns pending cash withdrawal.\", \"candidate_020640\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_461b3e\": \"The user's request concerns automatic top up.\", \"candidate_8182a8\": \"The user's request concerns exchange via app.\", \"candidate_4a741e\": \"The user's request concerns virtual card not working.\", \"candidate_30962e\": \"The user's request concerns order physical card.\", \"candidate_5bedcb\": \"The user's request concerns card acceptance.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0fa176\": 0.0, \"candidate_6754ac\": 0.0, \"candidate_020640\": 0.0, \"candidate_461b3e\": 0.0, \"candidate_8182a8\": 0.0, \"candidate_4a741e\": 1.0, \"candidate_30962e\": 0.0, \"candidate_5bedcb\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:81a6038bc6cabaaf64d87b95:choice", "group_id": "fresh12:banking77:81a6038bc6cabaaf64d87b95", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My disposable virtual card is not working when I try to use it at a point of sale transaction. What do I do now?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3fc31e\": \"The user's request concerns topping up by card.\", \"candidate_fa6f0f\": \"The user's request concerns virtual card not working.\", \"candidate_e1e0c3\": \"The user's request concerns activate my card.\", \"candidate_0bcde4\": \"The user's request concerns card about to expire.\", \"candidate_5c4d4e\": \"The user's request concerns getting spare card.\", \"candidate_d7b596\": \"The user's request concerns card arrival.\", \"candidate_bb2390\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_f1c21f\": \"The user's request concerns Refund not showing up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3fc31e\": 0.0, \"candidate_fa6f0f\": 1.0, \"candidate_e1e0c3\": 0.0, \"candidate_0bcde4\": 0.0, \"candidate_5c4d4e\": 0.0, \"candidate_d7b596\": 0.0, \"candidate_bb2390\": 0.0, \"candidate_f1c21f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3ea29d04def875de76df7091:choice", "group_id": "fresh12:banking77:3ea29d04def875de76df7091", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I cannot make transactions with my virtual card.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_063083\": \"The user's request concerns order physical card.\", \"candidate_6cc3e3\": \"The user's request concerns transfer into account.\", \"candidate_84970c\": \"The user's request concerns cancel transfer.\", \"candidate_6b9287\": \"The user's request concerns disposable card limits.\", \"candidate_9c888e\": \"The user's request concerns transaction charged twice.\", \"candidate_8cc4a4\": \"The user's request concerns virtual card not working.\", \"candidate_514127\": \"The user's request concerns atm support.\", \"candidate_3e8e69\": \"The user's request concerns supported cards and currencies.\"}}}}", "targets_json": "{\"decision\": {\"candidate_063083\": 0.0, \"candidate_6cc3e3\": 0.0, \"candidate_84970c\": 0.0, \"candidate_6b9287\": 0.0, \"candidate_9c888e\": 0.0, \"candidate_8cc4a4\": 1.0, \"candidate_514127\": 0.0, \"candidate_3e8e69\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a1bd56b5dd5dc32916d4bf1d:choice", "group_id": "fresh12:banking77:a1bd56b5dd5dc32916d4bf1d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What do I do if my virtual card won't work.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4680a3\": \"The user's request concerns beneficiary not allowed.\", \"candidate_5af833\": \"The user's request concerns virtual card not working.\", \"candidate_eeed70\": \"The user's request concerns top up failed.\", \"candidate_70ca44\": \"The user's request concerns pending cash withdrawal.\", \"candidate_5df5f6\": \"The user's request concerns card acceptance.\", \"candidate_eccb52\": \"The user's request concerns card payment fee charged.\", \"candidate_f3ea57\": \"The user's request concerns receiving money.\", \"candidate_e01535\": \"The user's request concerns terminate account.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4680a3\": 0.0, \"candidate_5af833\": 1.0, \"candidate_eeed70\": 0.0, \"candidate_70ca44\": 0.0, \"candidate_5df5f6\": 0.0, \"candidate_eccb52\": 0.0, \"candidate_f3ea57\": 0.0, \"candidate_e01535\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7c34dff72adf39c23df8a4ae:choice", "group_id": "fresh12:banking77:7c34dff72adf39c23df8a4ae", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The exchange rate for my withdrawal isn't correct and doesn't match the actual exchange rate.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ca21f4\": \"The user's request concerns declined card payment.\", \"candidate_91f862\": \"The user's request concerns getting spare card.\", \"candidate_094a07\": \"The user's request concerns atm support.\", \"candidate_d45c3f\": \"The user's request concerns card not working.\", \"candidate_820274\": \"The user's request concerns card swallowed.\", \"candidate_90813c\": \"The user's request concerns top up by card charge.\", \"candidate_fdd582\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_131147\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ca21f4\": 0.0, \"candidate_91f862\": 0.0, \"candidate_094a07\": 0.0, \"candidate_d45c3f\": 0.0, \"candidate_820274\": 0.0, \"candidate_90813c\": 0.0, \"candidate_fdd582\": 1.0, \"candidate_131147\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b2914affc7c0909e2c989b55:choice", "group_id": "fresh12:banking77:b2914affc7c0909e2c989b55", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"In receiving cash, the wrong exchange rate was used for my transaction.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_75e6ca\": \"The user's request concerns activate my card.\", \"candidate_c49d15\": \"The user's request concerns age limit.\", \"candidate_73bb41\": \"The user's request concerns getting virtual card.\", \"candidate_ab28b5\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_4a63a5\": \"The user's request concerns edit personal details.\", \"candidate_f05372\": \"The user's request concerns verify top up.\", \"candidate_8338ec\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_9d701a\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_75e6ca\": 0.0, \"candidate_c49d15\": 0.0, \"candidate_73bb41\": 0.0, \"candidate_ab28b5\": 1.0, \"candidate_4a63a5\": 0.0, \"candidate_f05372\": 0.0, \"candidate_8338ec\": 0.0, \"candidate_9d701a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ffd2639c378d0cabb493a948:choice", "group_id": "fresh12:banking77:ffd2639c378d0cabb493a948", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"The see the rate applied to my transaction was wrong at the atm.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_db8324\": \"The user's request concerns pin blocked.\", \"candidate_7436c9\": \"The user's request concerns order physical card.\", \"candidate_82e456\": \"The user's request concerns transfer into account.\", \"candidate_f65866\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_b52b4e\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_d49d5c\": \"The user's request concerns top up failed.\", \"candidate_32ba93\": \"The user's request concerns Refund not showing up.\", \"candidate_454a54\": \"The user's request concerns atm support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_db8324\": 0.0, \"candidate_7436c9\": 0.0, \"candidate_82e456\": 0.0, \"candidate_f65866\": 1.0, \"candidate_b52b4e\": 0.0, \"candidate_d49d5c\": 0.0, \"candidate_32ba93\": 0.0, \"candidate_454a54\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:aee4e8b7db18d7c1aeb586be:choice", "group_id": "fresh12:banking77:aee4e8b7db18d7c1aeb586be", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I made a withdraw from the ATM this past holiday and it seems like a was charged too much. If I would have   known about these charges I wouldn't have made a withdraw.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d2f955\": \"The user's request concerns top up limits.\", \"candidate_d33c04\": \"The user's request concerns country support.\", \"candidate_c19fed\": \"The user's request concerns exchange rate.\", \"candidate_f4d242\": \"The user's request concerns topping up by card.\", \"candidate_448990\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_06daee\": \"The user's request concerns declined card payment.\", \"candidate_e91462\": \"The user's request concerns compromised card.\", \"candidate_14a3d0\": \"The user's request concerns direct debit payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d2f955\": 0.0, \"candidate_d33c04\": 0.0, \"candidate_c19fed\": 0.0, \"candidate_f4d242\": 0.0, \"candidate_448990\": 1.0, \"candidate_06daee\": 0.0, \"candidate_e91462\": 0.0, \"candidate_14a3d0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:11d5e9790af8bc6b1cad3274:choice", "group_id": "fresh12:banking77:11d5e9790af8bc6b1cad3274", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I withdrew some cash out of the ATM over the holiday. It seems that I was charged some outrageous fees. I would not have done that had I been aware of these outrageous charges!\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_70a482\": \"The user's request concerns exchange via app.\", \"candidate_3c41fb\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_6fd097\": \"The user's request concerns card not working.\", \"candidate_39af00\": \"The user's request concerns card about to expire.\", \"candidate_a7a8d7\": \"The user's request concerns declined transfer.\", \"candidate_64fdc6\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_707e36\": \"The user's request concerns card payment fee charged.\", \"candidate_39162d\": \"The user's request concerns card delivery estimate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_70a482\": 0.0, \"candidate_3c41fb\": 0.0, \"candidate_6fd097\": 0.0, \"candidate_39af00\": 0.0, \"candidate_a7a8d7\": 0.0, \"candidate_64fdc6\": 1.0, \"candidate_707e36\": 0.0, \"candidate_39162d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:505941690167ca25a5438311:choice", "group_id": "fresh12:banking77:505941690167ca25a5438311", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Will there be additional costs of I make a withdrawal from a local ATM of British pounds? I need some cash to feel comfortable on the journey home\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7a0b47\": \"The user's request concerns unable to verify identity.\", \"candidate_4f9201\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_473eca\": \"The user's request concerns receiving money.\", \"candidate_4ab0a8\": \"The user's request concerns topping up by card.\", \"candidate_e9c95f\": \"The user's request concerns top up by cash or cheque.\", \"candidate_2d2c5e\": \"The user's request concerns why verify identity.\", \"candidate_3303e0\": \"The user's request concerns lost or stolen card.\", \"candidate_49220c\": \"The user's request concerns wrong exchange rate for cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7a0b47\": 0.0, \"candidate_4f9201\": 0.0, \"candidate_473eca\": 0.0, \"candidate_4ab0a8\": 0.0, \"candidate_e9c95f\": 0.0, \"candidate_2d2c5e\": 0.0, \"candidate_3303e0\": 0.0, \"candidate_49220c\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9b64a78a87387dcbcfd000f3:choice", "group_id": "fresh12:banking77:9b64a78a87387dcbcfd000f3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I just purchased a disposable virtual card. Am I able to use it for online purchases immediately?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9e4c07\": \"The user's request concerns passcode forgotten.\", \"candidate_ac0527\": \"The user's request concerns pending card payment.\", \"candidate_8b1f18\": \"The user's request concerns compromised card.\", \"candidate_dda055\": \"The user's request concerns wrong amount of cash received.\", \"candidate_c59680\": \"The user's request concerns get disposable virtual card.\", \"candidate_e7c3b5\": \"The user's request concerns transfer into account.\", \"candidate_754d4c\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_ee1fc7\": \"The user's request concerns age limit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9e4c07\": 0.0, \"candidate_ac0527\": 0.0, \"candidate_8b1f18\": 0.0, \"candidate_dda055\": 0.0, \"candidate_c59680\": 1.0, \"candidate_e7c3b5\": 0.0, \"candidate_754d4c\": 0.0, \"candidate_ee1fc7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c1c5a052dc7cac9682b7f7a1:choice", "group_id": "fresh12:banking77:c1c5a052dc7cac9682b7f7a1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do the disposable cards work?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_cc54b5\": \"The user's request concerns extra charge on statement.\", \"candidate_0f7012\": \"The user's request concerns declined card payment.\", \"candidate_38988d\": \"The user's request concerns cash withdrawal charge.\", \"candidate_af598b\": \"The user's request concerns order physical card.\", \"candidate_92d8cf\": \"The user's request concerns get disposable virtual card.\", \"candidate_34d546\": \"The user's request concerns request refund.\", \"candidate_f39c21\": \"The user's request concerns top up by cash or cheque.\", \"candidate_58a8aa\": \"The user's request concerns why verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_cc54b5\": 0.0, \"candidate_0f7012\": 0.0, \"candidate_38988d\": 0.0, \"candidate_af598b\": 0.0, \"candidate_92d8cf\": 1.0, \"candidate_34d546\": 0.0, \"candidate_f39c21\": 0.0, \"candidate_58a8aa\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1012f61d4842a00a8c592a46:choice", "group_id": "fresh12:banking77:1012f61d4842a00a8c592a46", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where's the best place to get a disposable virtual card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c0e61e\": \"The user's request concerns transfer not received by recipient.\", \"candidate_ed739a\": \"The user's request concerns pending top up.\", \"candidate_6ddd56\": \"The user's request concerns get disposable virtual card.\", \"candidate_6eaa09\": \"The user's request concerns pending card payment.\", \"candidate_39c529\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_93e6bc\": \"The user's request concerns wrong amount of cash received.\", \"candidate_07f2d7\": \"The user's request concerns getting spare card.\", \"candidate_24d058\": \"The user's request concerns visa or mastercard.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c0e61e\": 0.0, \"candidate_ed739a\": 0.0, \"candidate_6ddd56\": 1.0, \"candidate_6eaa09\": 0.0, \"candidate_39c529\": 0.0, \"candidate_93e6bc\": 0.0, \"candidate_07f2d7\": 0.0, \"candidate_24d058\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:2dbb76ff85f79e7a097c41d3:choice", "group_id": "fresh12:banking77:2dbb76ff85f79e7a097c41d3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need a disposable virtual card, how do I get one?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_36e1c6\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_9268b1\": \"The user's request concerns edit personal details.\", \"candidate_1af138\": \"The user's request concerns card swallowed.\", \"candidate_255f93\": \"The user's request concerns get disposable virtual card.\", \"candidate_55615c\": \"The user's request concerns pending top up.\", \"candidate_f0d165\": \"The user's request concerns top up limits.\", \"candidate_8361fc\": \"The user's request concerns transfer into account.\", \"candidate_cf24e6\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_36e1c6\": 0.0, \"candidate_9268b1\": 0.0, \"candidate_1af138\": 0.0, \"candidate_255f93\": 1.0, \"candidate_55615c\": 0.0, \"candidate_f0d165\": 0.0, \"candidate_8361fc\": 0.0, \"candidate_cf24e6\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:8b0985a69be2acb88e8d547e:choice", "group_id": "fresh12:banking77:8b0985a69be2acb88e8d547e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do disposable cards work?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_075256\": \"The user's request concerns card acceptance.\", \"candidate_ed6ae1\": \"The user's request concerns reverted card payment?.\", \"candidate_ee7f59\": \"The user's request concerns get disposable virtual card.\", \"candidate_5ac8b5\": \"The user's request concerns atm support.\", \"candidate_b308ca\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_2263c0\": \"The user's request concerns apple pay or google pay.\", \"candidate_2cc57f\": \"The user's request concerns card swallowed.\", \"candidate_3581ce\": \"The user's request concerns cancel transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_075256\": 0.0, \"candidate_ed6ae1\": 0.0, \"candidate_ee7f59\": 1.0, \"candidate_5ac8b5\": 0.0, \"candidate_b308ca\": 0.0, \"candidate_2263c0\": 0.0, \"candidate_2cc57f\": 0.0, \"candidate_3581ce\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:31b1e35150a2d003887a7530:choice", "group_id": "fresh12:banking77:31b1e35150a2d003887a7530", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where can I get a disposable virtual card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0674d2\": \"The user's request concerns card not working.\", \"candidate_6d06ea\": \"The user's request concerns exchange rate.\", \"candidate_4776ee\": \"The user's request concerns getting spare card.\", \"candidate_e51f1d\": \"The user's request concerns lost or stolen phone.\", \"candidate_d205a9\": \"The user's request concerns supported cards and currencies.\", \"candidate_eae307\": \"The user's request concerns card payment not recognised.\", \"candidate_5a85dd\": \"The user's request concerns card delivery estimate.\", \"candidate_0ef581\": \"The user's request concerns get disposable virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0674d2\": 0.0, \"candidate_6d06ea\": 0.0, \"candidate_4776ee\": 0.0, \"candidate_e51f1d\": 0.0, \"candidate_d205a9\": 0.0, \"candidate_eae307\": 0.0, \"candidate_5a85dd\": 0.0, \"candidate_0ef581\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a9f8de0d3609af0e067a1487:choice", "group_id": "fresh12:banking77:a9f8de0d3609af0e067a1487", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"So, I tried topping up my card for the today. Unfortunately, it failed. A couple of days back I used it and it works fine. Can you double check and tell me what's going on?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ff2437\": \"The user's request concerns verify top up.\", \"candidate_ffd188\": \"The user's request concerns Refund not showing up.\", \"candidate_be5fdb\": \"The user's request concerns why verify identity.\", \"candidate_d27d32\": \"The user's request concerns top up failed.\", \"candidate_fcf189\": \"The user's request concerns change pin.\", \"candidate_84e51f\": \"The user's request concerns transfer fee charged.\", \"candidate_2d3f03\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_40d3e7\": \"The user's request concerns top up by cash or cheque.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ff2437\": 0.0, \"candidate_ffd188\": 0.0, \"candidate_be5fdb\": 0.0, \"candidate_d27d32\": 1.0, \"candidate_fcf189\": 0.0, \"candidate_84e51f\": 0.0, \"candidate_2d3f03\": 0.0, \"candidate_40d3e7\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:318d57782c3aaca9d9a6b9a1:choice", "group_id": "fresh12:banking77:318d57782c3aaca9d9a6b9a1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Heyy I tried to top up with my card today but it failed! I just got the couple days before but it has worked last time. What's going on, can you double check please?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9fe107\": \"The user's request concerns card arrival.\", \"candidate_8edc2e\": \"The user's request concerns declined cash withdrawal.\", \"candidate_ddd12d\": \"The user's request concerns wrong amount of cash received.\", \"candidate_c20029\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_6764ad\": \"The user's request concerns top up failed.\", \"candidate_9ee969\": \"The user's request concerns getting virtual card.\", \"candidate_a14ddb\": \"The user's request concerns get disposable virtual card.\", \"candidate_7f825f\": \"The user's request concerns wrong exchange rate for cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9fe107\": 0.0, \"candidate_8edc2e\": 0.0, \"candidate_ddd12d\": 0.0, \"candidate_c20029\": 0.0, \"candidate_6764ad\": 1.0, \"candidate_9ee969\": 0.0, \"candidate_a14ddb\": 0.0, \"candidate_7f825f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b1ec0b12d0b8ae2efe506eea:choice", "group_id": "fresh12:banking77:b1ec0b12d0b8ae2efe506eea", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My top-up isn't working.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f876e3\": \"The user's request concerns disposable card limits.\", \"candidate_1f55b6\": \"The user's request concerns wrong amount of cash received.\", \"candidate_c9cf65\": \"The user's request concerns cancel transfer.\", \"candidate_18a2ca\": \"The user's request concerns top up limits.\", \"candidate_71c40e\": \"The user's request concerns top up failed.\", \"candidate_1c2db5\": \"The user's request concerns pin blocked.\", \"candidate_a38aa9\": \"The user's request concerns cash withdrawal charge.\", \"candidate_69fe4b\": \"The user's request concerns verify source of funds.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f876e3\": 0.0, \"candidate_1f55b6\": 0.0, \"candidate_c9cf65\": 0.0, \"candidate_18a2ca\": 0.0, \"candidate_71c40e\": 1.0, \"candidate_1c2db5\": 0.0, \"candidate_a38aa9\": 0.0, \"candidate_69fe4b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:89c6d122a619802a48d1372d:choice", "group_id": "fresh12:banking77:89c6d122a619802a48d1372d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I topped up but the app did not accept it.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7888cc\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_42d815\": \"The user's request concerns declined cash withdrawal.\", \"candidate_5c2ef3\": \"The user's request concerns top up failed.\", \"candidate_f35707\": \"The user's request concerns card about to expire.\", \"candidate_42541b\": \"The user's request concerns why verify identity.\", \"candidate_5c164a\": \"The user's request concerns exchange rate.\", \"candidate_cd226b\": \"The user's request concerns visa or mastercard.\", \"candidate_13d6e0\": \"The user's request concerns get physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7888cc\": 0.0, \"candidate_42d815\": 0.0, \"candidate_5c2ef3\": 1.0, \"candidate_f35707\": 0.0, \"candidate_42541b\": 0.0, \"candidate_5c164a\": 0.0, \"candidate_cd226b\": 0.0, \"candidate_13d6e0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:18aba6b60ee14bc6c26d8d17:choice", "group_id": "fresh12:banking77:18aba6b60ee14bc6c26d8d17", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Problems with top up\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_225616\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_b42fa3\": \"The user's request concerns exchange via app.\", \"candidate_ac8f9a\": \"The user's request concerns card about to expire.\", \"candidate_55ae16\": \"The user's request concerns cancel transfer.\", \"candidate_f1616b\": \"The user's request concerns top up failed.\", \"candidate_9f01b2\": \"The user's request concerns terminate account.\", \"candidate_160d6b\": \"The user's request concerns declined card payment.\", \"candidate_8c3994\": \"The user's request concerns top up by card charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_225616\": 0.0, \"candidate_b42fa3\": 0.0, \"candidate_ac8f9a\": 0.0, \"candidate_55ae16\": 0.0, \"candidate_f1616b\": 1.0, \"candidate_9f01b2\": 0.0, \"candidate_160d6b\": 0.0, \"candidate_8c3994\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:92d1941adfd1e47c0ab76092:choice", "group_id": "fresh12:banking77:92d1941adfd1e47c0ab76092", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My top up didn't work.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_34d2a5\": \"The user's request concerns top up failed.\", \"candidate_9740ed\": \"The user's request concerns lost or stolen card.\", \"candidate_7c0ad3\": \"The user's request concerns card delivery estimate.\", \"candidate_ae2094\": \"The user's request concerns card arrival.\", \"candidate_b2ad83\": \"The user's request concerns pending card payment.\", \"candidate_57e0d4\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_43edcf\": \"The user's request concerns top up by card charge.\", \"candidate_280206\": \"The user's request concerns wrong amount of cash received.\"}}}}", "targets_json": "{\"decision\": {\"candidate_34d2a5\": 1.0, \"candidate_9740ed\": 0.0, \"candidate_7c0ad3\": 0.0, \"candidate_ae2094\": 0.0, \"candidate_b2ad83\": 0.0, \"candidate_57e0d4\": 0.0, \"candidate_43edcf\": 0.0, \"candidate_280206\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:355754428a7729e8b7a978af:choice", "group_id": "fresh12:banking77:355754428a7729e8b7a978af", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I made a transfer a few hours ago from a UK banking account.  I do not yet see the transfer.  Please check on this for me.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_bbf182\": \"The user's request concerns top up limits.\", \"candidate_eb9454\": \"The user's request concerns unable to verify identity.\", \"candidate_c92604\": \"The user's request concerns order physical card.\", \"candidate_47d9f2\": \"The user's request concerns card linking.\", \"candidate_5c2e2c\": \"The user's request concerns activate my card.\", \"candidate_9368fb\": \"The user's request concerns edit personal details.\", \"candidate_23f210\": \"The user's request concerns get disposable virtual card.\", \"candidate_5e7539\": \"The user's request concerns balance not updated after bank transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_bbf182\": 0.0, \"candidate_eb9454\": 0.0, \"candidate_c92604\": 0.0, \"candidate_47d9f2\": 0.0, \"candidate_5c2e2c\": 0.0, \"candidate_9368fb\": 0.0, \"candidate_23f210\": 0.0, \"candidate_5e7539\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6e80d7b61e9161840cd86b49:choice", "group_id": "fresh12:banking77:6e80d7b61e9161840cd86b49", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Where is my money? I transfered it and it isn't in my account.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_e53d9f\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_75dc5a\": \"The user's request concerns card about to expire.\", \"candidate_8ae0bb\": \"The user's request concerns pending top up.\", \"candidate_a022d1\": \"The user's request concerns declined card payment.\", \"candidate_7baeab\": \"The user's request concerns beneficiary not allowed.\", \"candidate_e9179e\": \"The user's request concerns change pin.\", \"candidate_dbed82\": \"The user's request concerns order physical card.\", \"candidate_26a990\": \"The user's request concerns wrong exchange rate for cash withdrawal.\"}}}}", "targets_json": "{\"decision\": {\"candidate_e53d9f\": 1.0, \"candidate_75dc5a\": 0.0, \"candidate_8ae0bb\": 0.0, \"candidate_a022d1\": 0.0, \"candidate_7baeab\": 0.0, \"candidate_e9179e\": 0.0, \"candidate_dbed82\": 0.0, \"candidate_26a990\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:36c405d92885eec68f9417f7:choice", "group_id": "fresh12:banking77:36c405d92885eec68f9417f7", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Hey, I try to make a bank transfer from my UK account a few hours ago, but it haven't show up yet. Please check my account to make sure everything alright with it?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d7e101\": \"The user's request concerns card not working.\", \"candidate_32d3d3\": \"The user's request concerns atm support.\", \"candidate_0a09f3\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_ee1b46\": \"The user's request concerns getting spare card.\", \"candidate_b3700a\": \"The user's request concerns top up reverted.\", \"candidate_fda550\": \"The user's request concerns change pin.\", \"candidate_30e6cc\": \"The user's request concerns automatic top up.\", \"candidate_6344ca\": \"The user's request concerns age limit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d7e101\": 0.0, \"candidate_32d3d3\": 0.0, \"candidate_0a09f3\": 1.0, \"candidate_ee1b46\": 0.0, \"candidate_b3700a\": 0.0, \"candidate_fda550\": 0.0, \"candidate_30e6cc\": 0.0, \"candidate_6344ca\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4d19b2cbd640283d1f0031b7:choice", "group_id": "fresh12:banking77:4d19b2cbd640283d1f0031b7", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My transfer is pending.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_d92347\": \"The user's request concerns transfer not received by recipient.\", \"candidate_c8502e\": \"The user's request concerns lost or stolen card.\", \"candidate_f1d02d\": \"The user's request concerns age limit.\", \"candidate_190ece\": \"The user's request concerns cash withdrawal charge.\", \"candidate_8a0544\": \"The user's request concerns card payment fee charged.\", \"candidate_7e93c0\": \"The user's request concerns fiat currency support.\", \"candidate_3cd053\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_2dbe62\": \"The user's request concerns card arrival.\"}}}}", "targets_json": "{\"decision\": {\"candidate_d92347\": 0.0, \"candidate_c8502e\": 0.0, \"candidate_f1d02d\": 0.0, \"candidate_190ece\": 0.0, \"candidate_8a0544\": 0.0, \"candidate_7e93c0\": 0.0, \"candidate_3cd053\": 1.0, \"candidate_2dbe62\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f9dcb11e0faacf2c06de5ec2:choice", "group_id": "fresh12:banking77:f9dcb11e0faacf2c06de5ec2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why don't I have my transfer?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2fd6fd\": \"The user's request concerns beneficiary not allowed.\", \"candidate_575e65\": \"The user's request concerns transfer not received by recipient.\", \"candidate_78a1cf\": \"The user's request concerns why verify identity.\", \"candidate_0d96e5\": \"The user's request concerns transaction charged twice.\", \"candidate_1d4857\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_0aff92\": \"The user's request concerns activate my card.\", \"candidate_6900e3\": \"The user's request concerns get disposable virtual card.\", \"candidate_99b473\": \"The user's request concerns balance not updated after bank transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2fd6fd\": 0.0, \"candidate_575e65\": 0.0, \"candidate_78a1cf\": 0.0, \"candidate_0d96e5\": 0.0, \"candidate_1d4857\": 0.0, \"candidate_0aff92\": 0.0, \"candidate_6900e3\": 0.0, \"candidate_99b473\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3f0a6fe2c98cc43b845d2f50:choice", "group_id": "fresh12:banking77:3f0a6fe2c98cc43b845d2f50", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My account balance has not gone up even though I just transferred money into it\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9467fc\": \"The user's request concerns top up by cash or cheque.\", \"candidate_28684c\": \"The user's request concerns exchange rate.\", \"candidate_e3e26c\": \"The user's request concerns compromised card.\", \"candidate_3ed5b9\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_341aa7\": \"The user's request concerns pin blocked.\", \"candidate_d67251\": \"The user's request concerns cancel transfer.\", \"candidate_9bb2f9\": \"The user's request concerns change pin.\", \"candidate_3455de\": \"The user's request concerns visa or mastercard.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9467fc\": 0.0, \"candidate_28684c\": 0.0, \"candidate_e3e26c\": 0.0, \"candidate_3ed5b9\": 1.0, \"candidate_341aa7\": 0.0, \"candidate_d67251\": 0.0, \"candidate_9bb2f9\": 0.0, \"candidate_3455de\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5cf2d42cdc47124ee54e791a:choice", "group_id": "fresh12:banking77:5cf2d42cdc47124ee54e791a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I just lost my wallet and I see that they are already withdrawing money from my account. How can I stop this?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_aaa946\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_009a7f\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_54fcfb\": \"The user's request concerns transfer timing.\", \"candidate_1d6bba\": \"The user's request concerns country support.\", \"candidate_9a2c75\": \"The user's request concerns top up by cash or cheque.\", \"candidate_9befb4\": \"The user's request concerns card payment not recognised.\", \"candidate_9ccea3\": \"The user's request concerns transfer into account.\", \"candidate_77539a\": \"The user's request concerns compromised card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_aaa946\": 0.0, \"candidate_009a7f\": 1.0, \"candidate_54fcfb\": 0.0, \"candidate_1d6bba\": 0.0, \"candidate_9a2c75\": 0.0, \"candidate_9befb4\": 0.0, \"candidate_9ccea3\": 0.0, \"candidate_77539a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b3be05b3453cf50ccb073cb2:choice", "group_id": "fresh12:banking77:b3be05b3453cf50ccb073cb2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is a odd withdrawal on my account.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f76646\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_aa8a4a\": \"The user's request concerns verify top up.\", \"candidate_33dc51\": \"The user's request concerns getting spare card.\", \"candidate_c8247e\": \"The user's request concerns transfer not received by recipient.\", \"candidate_fcf07b\": \"The user's request concerns verify my identity.\", \"candidate_f51bae\": \"The user's request concerns top up by cash or cheque.\", \"candidate_801159\": \"The user's request concerns card payment fee charged.\", \"candidate_57e7ba\": \"The user's request concerns contactless not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f76646\": 1.0, \"candidate_aa8a4a\": 0.0, \"candidate_33dc51\": 0.0, \"candidate_c8247e\": 0.0, \"candidate_fcf07b\": 0.0, \"candidate_f51bae\": 0.0, \"candidate_801159\": 0.0, \"candidate_57e7ba\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:b470654a55f8673219cc3be4:choice", "group_id": "fresh12:banking77:b470654a55f8673219cc3be4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is an unauthorized cash withdrawal from my account for 500£.  I definitely didn't make this withdrawal.  I need your help.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0da7bb\": \"The user's request concerns verify my identity.\", \"candidate_9558e4\": \"The user's request concerns top up by cash or cheque.\", \"candidate_953d05\": \"The user's request concerns reverted card payment?.\", \"candidate_5b9437\": \"The user's request concerns age limit.\", \"candidate_803b72\": \"The user's request concerns pending top up.\", \"candidate_8ea4e0\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_222c26\": \"The user's request concerns pending card payment.\", \"candidate_12e2b0\": \"The user's request concerns get physical card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0da7bb\": 0.0, \"candidate_9558e4\": 0.0, \"candidate_953d05\": 0.0, \"candidate_5b9437\": 0.0, \"candidate_803b72\": 0.0, \"candidate_8ea4e0\": 1.0, \"candidate_222c26\": 0.0, \"candidate_12e2b0\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:67694b99fc73cda966822a5b:choice", "group_id": "fresh12:banking77:67694b99fc73cda966822a5b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My app says that I received cash from an ATM and I didn't.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_851eec\": \"The user's request concerns edit personal details.\", \"candidate_cdf1a6\": \"The user's request concerns pending cash withdrawal.\", \"candidate_21f05b\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_455cb4\": \"The user's request concerns apple pay or google pay.\", \"candidate_3de282\": \"The user's request concerns exchange rate.\", \"candidate_7dafd1\": \"The user's request concerns card linking.\", \"candidate_3f3907\": \"The user's request concerns terminate account.\", \"candidate_a55a90\": \"The user's request concerns supported cards and currencies.\"}}}}", "targets_json": "{\"decision\": {\"candidate_851eec\": 0.0, \"candidate_cdf1a6\": 0.0, \"candidate_21f05b\": 1.0, \"candidate_455cb4\": 0.0, \"candidate_3de282\": 0.0, \"candidate_7dafd1\": 0.0, \"candidate_3f3907\": 0.0, \"candidate_a55a90\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1aae622e66cedccdd36356a3:choice", "group_id": "fresh12:banking77:1aae622e66cedccdd36356a3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is a cash withdrawal transaction that I am unsure of.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4302a4\": \"The user's request concerns card linking.\", \"candidate_70035e\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_4bd8d3\": \"The user's request concerns extra charge on statement.\", \"candidate_4444dd\": \"The user's request concerns transfer not received by recipient.\", \"candidate_ee90e3\": \"The user's request concerns top up limits.\", \"candidate_0d6666\": \"The user's request concerns card swallowed.\", \"candidate_d71cbc\": \"The user's request concerns Refund not showing up.\", \"candidate_5014af\": \"The user's request concerns exchange charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4302a4\": 0.0, \"candidate_70035e\": 1.0, \"candidate_4bd8d3\": 0.0, \"candidate_4444dd\": 0.0, \"candidate_ee90e3\": 0.0, \"candidate_0d6666\": 0.0, \"candidate_d71cbc\": 0.0, \"candidate_5014af\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:f2e5727312852409d9084f37:choice", "group_id": "fresh12:banking77:f2e5727312852409d9084f37", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"There is an ATM withdrawl in the app I don't recognize.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c02e1c\": \"The user's request concerns cash withdrawal charge.\", \"candidate_ebe7bf\": \"The user's request concerns lost or stolen phone.\", \"candidate_47ce9f\": \"The user's request concerns card not working.\", \"candidate_2d44b5\": \"The user's request concerns request refund.\", \"candidate_9fd09e\": \"The user's request concerns get physical card.\", \"candidate_84b2d9\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_630a42\": \"The user's request concerns age limit.\", \"candidate_4b3272\": \"The user's request concerns direct debit payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c02e1c\": 0.0, \"candidate_ebe7bf\": 0.0, \"candidate_47ce9f\": 0.0, \"candidate_2d44b5\": 0.0, \"candidate_9fd09e\": 0.0, \"candidate_84b2d9\": 1.0, \"candidate_630a42\": 0.0, \"candidate_4b3272\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:fa2c9d1e5fc1f4aa26103cb9:choice", "group_id": "fresh12:banking77:fa2c9d1e5fc1f4aa26103cb9", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How much more do I have to pay to exchange currencies?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7ace38\": \"The user's request concerns age limit.\", \"candidate_cbda1e\": \"The user's request concerns top up by card charge.\", \"candidate_49b8c9\": \"The user's request concerns exchange via app.\", \"candidate_538de8\": \"The user's request concerns getting virtual card.\", \"candidate_63d875\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_40d4a2\": \"The user's request concerns exchange charge.\", \"candidate_c9b6f7\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_2adea8\": \"The user's request concerns lost or stolen phone.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7ace38\": 0.0, \"candidate_cbda1e\": 0.0, \"candidate_49b8c9\": 0.0, \"candidate_538de8\": 0.0, \"candidate_63d875\": 0.0, \"candidate_40d4a2\": 1.0, \"candidate_c9b6f7\": 0.0, \"candidate_2adea8\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a841706a056db88b82e08e93:choice", "group_id": "fresh12:banking77:a841706a056db88b82e08e93", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Are there any extra hidden fees for exchanging currencies?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_65b184\": \"The user's request concerns exchange charge.\", \"candidate_86e729\": \"The user's request concerns automatic top up.\", \"candidate_c6588c\": \"The user's request concerns disposable card limits.\", \"candidate_7d63f0\": \"The user's request concerns card payment fee charged.\", \"candidate_0daa2a\": \"The user's request concerns exchange rate.\", \"candidate_ce316e\": \"The user's request concerns topping up by card.\", \"candidate_9c765c\": \"The user's request concerns top up reverted.\", \"candidate_8b1b59\": \"The user's request concerns balance not updated after cheque or cash deposit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_65b184\": 1.0, \"candidate_86e729\": 0.0, \"candidate_c6588c\": 0.0, \"candidate_7d63f0\": 0.0, \"candidate_0daa2a\": 0.0, \"candidate_ce316e\": 0.0, \"candidate_9c765c\": 0.0, \"candidate_8b1b59\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:089746b47c75a4a740598bdd:choice", "group_id": "fresh12:banking77:089746b47c75a4a740598bdd", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Would I be charged for exchanging currencies?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c2ec1b\": \"The user's request concerns cash withdrawal charge.\", \"candidate_56f2da\": \"The user's request concerns request refund.\", \"candidate_a0c8c1\": \"The user's request concerns exchange charge.\", \"candidate_a4b7af\": \"The user's request concerns activate my card.\", \"candidate_1ae90c\": \"The user's request concerns pending cash withdrawal.\", \"candidate_06d96c\": \"The user's request concerns card arrival.\", \"candidate_6a4053\": \"The user's request concerns edit personal details.\", \"candidate_53b8a1\": \"The user's request concerns terminate account.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c2ec1b\": 0.0, \"candidate_56f2da\": 0.0, \"candidate_a0c8c1\": 1.0, \"candidate_a4b7af\": 0.0, \"candidate_1ae90c\": 0.0, \"candidate_06d96c\": 0.0, \"candidate_6a4053\": 0.0, \"candidate_53b8a1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1bd42ff1e4b1ff48d3b6ea5c:choice", "group_id": "fresh12:banking77:1bd42ff1e4b1ff48d3b6ea5c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What are your currency exchange fees?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_68eaa3\": \"The user's request concerns supported cards and currencies.\", \"candidate_5dbb37\": \"The user's request concerns cash withdrawal charge.\", \"candidate_dfbe95\": \"The user's request concerns pin blocked.\", \"candidate_e38491\": \"The user's request concerns failed transfer.\", \"candidate_a195f0\": \"The user's request concerns transfer not received by recipient.\", \"candidate_e2bf43\": \"The user's request concerns pending top up.\", \"candidate_fc4a9f\": \"The user's request concerns exchange charge.\", \"candidate_b2c327\": \"The user's request concerns card acceptance.\"}}}}", "targets_json": "{\"decision\": {\"candidate_68eaa3\": 0.0, \"candidate_5dbb37\": 0.0, \"candidate_dfbe95\": 0.0, \"candidate_e38491\": 0.0, \"candidate_a195f0\": 0.0, \"candidate_e2bf43\": 0.0, \"candidate_fc4a9f\": 1.0, \"candidate_b2c327\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:322105fbe6c6180723ce5f9c:choice", "group_id": "fresh12:banking77:322105fbe6c6180723ce5f9c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there a discount for frequently exchanging currencies?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0b781e\": \"The user's request concerns get disposable virtual card.\", \"candidate_556733\": \"The user's request concerns atm support.\", \"candidate_abbdaf\": \"The user's request concerns lost or stolen phone.\", \"candidate_12020f\": \"The user's request concerns card arrival.\", \"candidate_da0af2\": \"The user's request concerns transfer timing.\", \"candidate_5816e7\": \"The user's request concerns verify top up.\", \"candidate_620e58\": \"The user's request concerns exchange charge.\", \"candidate_ed9999\": \"The user's request concerns card delivery estimate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0b781e\": 0.0, \"candidate_556733\": 0.0, \"candidate_abbdaf\": 0.0, \"candidate_12020f\": 0.0, \"candidate_da0af2\": 0.0, \"candidate_5816e7\": 0.0, \"candidate_620e58\": 1.0, \"candidate_ed9999\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:48c50fae83b7f1ac3c3f6fd2:choice", "group_id": "fresh12:banking77:48c50fae83b7f1ac3c3f6fd2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"If I want to exchange currency, will there be extras?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_86c6f8\": \"The user's request concerns transfer not received by recipient.\", \"candidate_b7b90e\": \"The user's request concerns pending cash withdrawal.\", \"candidate_ed0372\": \"The user's request concerns exchange charge.\", \"candidate_a4d455\": \"The user's request concerns exchange via app.\", \"candidate_a37b45\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_c0e665\": \"The user's request concerns pending card payment.\", \"candidate_9c4b0c\": \"The user's request concerns passcode forgotten.\", \"candidate_9b9fc1\": \"The user's request concerns transfer timing.\"}}}}", "targets_json": "{\"decision\": {\"candidate_86c6f8\": 0.0, \"candidate_b7b90e\": 0.0, \"candidate_ed0372\": 1.0, \"candidate_a4d455\": 0.0, \"candidate_a37b45\": 0.0, \"candidate_c0e665\": 0.0, \"candidate_9c4b0c\": 0.0, \"candidate_9b9fc1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4eb29d09fa74eaee8e7e8382:choice", "group_id": "fresh12:banking77:4eb29d09fa74eaee8e7e8382", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Are there any fees if I use a European bank card for a top up?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2f40db\": \"The user's request concerns declined transfer.\", \"candidate_c6476d\": \"The user's request concerns transfer timing.\", \"candidate_3c531c\": \"The user's request concerns top up by card charge.\", \"candidate_710296\": \"The user's request concerns receiving money.\", \"candidate_9cfecc\": \"The user's request concerns card not working.\", \"candidate_6254ac\": \"The user's request concerns atm support.\", \"candidate_4dd8db\": \"The user's request concerns apple pay or google pay.\", \"candidate_5f3c35\": \"The user's request concerns transfer into account.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2f40db\": 0.0, \"candidate_c6476d\": 0.0, \"candidate_3c531c\": 1.0, \"candidate_710296\": 0.0, \"candidate_9cfecc\": 0.0, \"candidate_6254ac\": 0.0, \"candidate_4dd8db\": 0.0, \"candidate_5f3c35\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:edcff30be32c774c2b0bd9a4:choice", "group_id": "fresh12:banking77:edcff30be32c774c2b0bd9a4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is their a fee for top ups?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c672b0\": \"The user's request concerns pending cash withdrawal.\", \"candidate_53c68e\": \"The user's request concerns verify my identity.\", \"candidate_0c0be4\": \"The user's request concerns card swallowed.\", \"candidate_57b462\": \"The user's request concerns get physical card.\", \"candidate_ae994e\": \"The user's request concerns request refund.\", \"candidate_4db040\": \"The user's request concerns apple pay or google pay.\", \"candidate_cda002\": \"The user's request concerns top up by card charge.\", \"candidate_f331fe\": \"The user's request concerns transfer timing.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c672b0\": 0.0, \"candidate_53c68e\": 0.0, \"candidate_0c0be4\": 0.0, \"candidate_57b462\": 0.0, \"candidate_ae994e\": 0.0, \"candidate_4db040\": 0.0, \"candidate_cda002\": 1.0, \"candidate_f331fe\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1eda2a930bae3f1c9aa74e14:choice", "group_id": "fresh12:banking77:1eda2a930bae3f1c9aa74e14", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Will I be charged for topping up by card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_069f2b\": \"The user's request concerns change pin.\", \"candidate_bbb7ed\": \"The user's request concerns verify my identity.\", \"candidate_2e4489\": \"The user's request concerns fiat currency support.\", \"candidate_86d72f\": \"The user's request concerns receiving money.\", \"candidate_769d6e\": \"The user's request concerns getting virtual card.\", \"candidate_9b548e\": \"The user's request concerns top up by card charge.\", \"candidate_bde5b4\": \"The user's request concerns request refund.\", \"candidate_7f26bd\": \"The user's request concerns exchange via app.\"}}}}", "targets_json": "{\"decision\": {\"candidate_069f2b\": 0.0, \"candidate_bbb7ed\": 0.0, \"candidate_2e4489\": 0.0, \"candidate_86d72f\": 0.0, \"candidate_769d6e\": 0.0, \"candidate_9b548e\": 1.0, \"candidate_bde5b4\": 0.0, \"candidate_7f26bd\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6a011406772ac8b204d29831:choice", "group_id": "fresh12:banking77:6a011406772ac8b204d29831", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Are there any extra fees for adding money into the international card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9c7e4e\": \"The user's request concerns terminate account.\", \"candidate_72c8f9\": \"The user's request concerns pending card payment.\", \"candidate_7482d2\": \"The user's request concerns compromised card.\", \"candidate_902feb\": \"The user's request concerns top up by card charge.\", \"candidate_815933\": \"The user's request concerns top up by cash or cheque.\", \"candidate_074524\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_efea73\": \"The user's request concerns unable to verify identity.\", \"candidate_dcda5b\": \"The user's request concerns cash withdrawal not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9c7e4e\": 0.0, \"candidate_72c8f9\": 0.0, \"candidate_7482d2\": 0.0, \"candidate_902feb\": 1.0, \"candidate_815933\": 0.0, \"candidate_074524\": 0.0, \"candidate_efea73\": 0.0, \"candidate_dcda5b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c0206b77c5fa07d72d28252d:choice", "group_id": "fresh12:banking77:c0206b77c5fa07d72d28252d", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is it okay to use a bank card to top up\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_42d5e2\": \"The user's request concerns top up by card charge.\", \"candidate_eca82e\": \"The user's request concerns unable to verify identity.\", \"candidate_b04957\": \"The user's request concerns change pin.\", \"candidate_511c02\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_14a22c\": \"The user's request concerns pending card payment.\", \"candidate_755252\": \"The user's request concerns why verify identity.\", \"candidate_346527\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_6c9fe1\": \"The user's request concerns card delivery estimate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_42d5e2\": 1.0, \"candidate_eca82e\": 0.0, \"candidate_b04957\": 0.0, \"candidate_511c02\": 0.0, \"candidate_14a22c\": 0.0, \"candidate_755252\": 0.0, \"candidate_346527\": 0.0, \"candidate_6c9fe1\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a8b6a8ddab1b2273c9c0a048:choice", "group_id": "fresh12:banking77:a8b6a8ddab1b2273c9c0a048", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Will I get charged for topping off my card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_09327a\": \"The user's request concerns apple pay or google pay.\", \"candidate_4a6d3a\": \"The user's request concerns country support.\", \"candidate_c48711\": \"The user's request concerns card about to expire.\", \"candidate_ed1675\": \"The user's request concerns card linking.\", \"candidate_457fe8\": \"The user's request concerns top up by card charge.\", \"candidate_1c71a6\": \"The user's request concerns transfer fee charged.\", \"candidate_b742b9\": \"The user's request concerns atm support.\", \"candidate_e7db2c\": \"The user's request concerns pending transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_09327a\": 0.0, \"candidate_4a6d3a\": 0.0, \"candidate_c48711\": 0.0, \"candidate_ed1675\": 0.0, \"candidate_457fe8\": 1.0, \"candidate_1c71a6\": 0.0, \"candidate_b742b9\": 0.0, \"candidate_e7db2c\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0edf530c7dd3461d95392f7c:choice", "group_id": "fresh12:banking77:0edf530c7dd3461d95392f7c", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need to activate my card can you do that now?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a74efa\": \"The user's request concerns why verify identity.\", \"candidate_2de2b4\": \"The user's request concerns get disposable virtual card.\", \"candidate_8c52d5\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_bfabce\": \"The user's request concerns age limit.\", \"candidate_5277bf\": \"The user's request concerns pending card payment.\", \"candidate_6dc850\": \"The user's request concerns activate my card.\", \"candidate_2d2cdc\": \"The user's request concerns disposable card limits.\", \"candidate_8541d4\": \"The user's request concerns supported cards and currencies.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a74efa\": 0.0, \"candidate_2de2b4\": 0.0, \"candidate_8c52d5\": 0.0, \"candidate_bfabce\": 0.0, \"candidate_5277bf\": 0.0, \"candidate_6dc850\": 1.0, \"candidate_2d2cdc\": 0.0, \"candidate_8541d4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:dc627202b983f93ede66e367:choice", "group_id": "fresh12:banking77:dc627202b983f93ede66e367", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I tried activating my bill of fare and it didn't workplace\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_1b80f9\": \"The user's request concerns supported cards and currencies.\", \"candidate_51b430\": \"The user's request concerns passcode forgotten.\", \"candidate_e3ef40\": \"The user's request concerns apple pay or google pay.\", \"candidate_6c5367\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_3aaa31\": \"The user's request concerns activate my card.\", \"candidate_7be315\": \"The user's request concerns Refund not showing up.\", \"candidate_86e1af\": \"The user's request concerns verify source of funds.\", \"candidate_058f77\": \"The user's request concerns card acceptance.\"}}}}", "targets_json": "{\"decision\": {\"candidate_1b80f9\": 0.0, \"candidate_51b430\": 0.0, \"candidate_e3ef40\": 0.0, \"candidate_6c5367\": 0.0, \"candidate_3aaa31\": 1.0, \"candidate_7be315\": 0.0, \"candidate_86e1af\": 0.0, \"candidate_058f77\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:dedaf72e76a36df0d71ef324:choice", "group_id": "fresh12:banking77:dedaf72e76a36df0d71ef324", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I got my new card. How do I activate it?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_723fcd\": \"The user's request concerns card delivery estimate.\", \"candidate_7a4f48\": \"The user's request concerns age limit.\", \"candidate_407964\": \"The user's request concerns pending top up.\", \"candidate_d7c6aa\": \"The user's request concerns activate my card.\", \"candidate_2f04de\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_0ee3d4\": \"The user's request concerns getting virtual card.\", \"candidate_7755d6\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_98bc57\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_723fcd\": 0.0, \"candidate_7a4f48\": 0.0, \"candidate_407964\": 0.0, \"candidate_d7c6aa\": 1.0, \"candidate_2f04de\": 0.0, \"candidate_0ee3d4\": 0.0, \"candidate_7755d6\": 0.0, \"candidate_98bc57\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a5af54fe4ecf7813bdb0b7f2:choice", "group_id": "fresh12:banking77:a5af54fe4ecf7813bdb0b7f2", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Activating my card\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_3fe36d\": \"The user's request concerns visa or mastercard.\", \"candidate_59b725\": \"The user's request concerns lost or stolen phone.\", \"candidate_07b033\": \"The user's request concerns beneficiary not allowed.\", \"candidate_685452\": \"The user's request concerns pending top up.\", \"candidate_c1068f\": \"The user's request concerns activate my card.\", \"candidate_503d94\": \"The user's request concerns transfer timing.\", \"candidate_8bce2a\": \"The user's request concerns edit personal details.\", \"candidate_c3f22f\": \"The user's request concerns card delivery estimate.\"}}}}", "targets_json": "{\"decision\": {\"candidate_3fe36d\": 0.0, \"candidate_59b725\": 0.0, \"candidate_07b033\": 0.0, \"candidate_685452\": 0.0, \"candidate_c1068f\": 1.0, \"candidate_503d94\": 0.0, \"candidate_8bce2a\": 0.0, \"candidate_c3f22f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:9e41c77d090fe3225b1d344f:choice", "group_id": "fresh12:banking77:9e41c77d090fe3225b1d344f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I have a new card and I want to activate it\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a68cc7\": \"The user's request concerns get physical card.\", \"candidate_2665ad\": \"The user's request concerns automatic top up.\", \"candidate_767a78\": \"The user's request concerns contactless not working.\", \"candidate_49241a\": \"The user's request concerns fiat currency support.\", \"candidate_33c9b2\": \"The user's request concerns request refund.\", \"candidate_351126\": \"The user's request concerns visa or mastercard.\", \"candidate_fcd97f\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_f1d7e3\": \"The user's request concerns activate my card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a68cc7\": 0.0, \"candidate_2665ad\": 0.0, \"candidate_767a78\": 0.0, \"candidate_49241a\": 0.0, \"candidate_33c9b2\": 0.0, \"candidate_351126\": 0.0, \"candidate_fcd97f\": 0.0, \"candidate_f1d7e3\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7fc5d2a437ce1ad2ba571b3a:choice", "group_id": "fresh12:banking77:7fc5d2a437ce1ad2ba571b3a", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How long will it take to activate my new card?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_0ccf0f\": \"The user's request concerns transfer into account.\", \"candidate_8bac96\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_3d108c\": \"The user's request concerns card delivery estimate.\", \"candidate_1ef090\": \"The user's request concerns exchange via app.\", \"candidate_df5f15\": \"The user's request concerns activate my card.\", \"candidate_3c77c7\": \"The user's request concerns verify top up.\", \"candidate_4aa27e\": \"The user's request concerns age limit.\", \"candidate_933c12\": \"The user's request concerns cash withdrawal charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_0ccf0f\": 0.0, \"candidate_8bac96\": 0.0, \"candidate_3d108c\": 0.0, \"candidate_1ef090\": 0.0, \"candidate_df5f15\": 1.0, \"candidate_3c77c7\": 0.0, \"candidate_4aa27e\": 0.0, \"candidate_933c12\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4537243ae835b1cde0452103:choice", "group_id": "fresh12:banking77:4537243ae835b1cde0452103", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"It looks like I was charged a withdrawal fee for going to my ATM. Why is that because I haven't been charged previously for doing so?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4af334\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_c7c0d3\": \"The user's request concerns cash withdrawal charge.\", \"candidate_1b2fbe\": \"The user's request concerns transfer fee charged.\", \"candidate_62df09\": \"The user's request concerns reverted card payment?.\", \"candidate_ed6307\": \"The user's request concerns card acceptance.\", \"candidate_78d297\": \"The user's request concerns wrong exchange rate for cash withdrawal.\", \"candidate_878b12\": \"The user's request concerns get physical card.\", \"candidate_202374\": \"The user's request concerns transfer into account.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4af334\": 0.0, \"candidate_c7c0d3\": 1.0, \"candidate_1b2fbe\": 0.0, \"candidate_62df09\": 0.0, \"candidate_ed6307\": 0.0, \"candidate_78d297\": 0.0, \"candidate_878b12\": 0.0, \"candidate_202374\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:d4a290b56ac7ac6f81422d2b:choice", "group_id": "fresh12:banking77:d4a290b56ac7ac6f81422d2b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why am I being charged when I withdraw cash?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_132d93\": \"The user's request concerns topping up by card.\", \"candidate_748ae1\": \"The user's request concerns exchange via app.\", \"candidate_542ef7\": \"The user's request concerns transfer not received by recipient.\", \"candidate_a28814\": \"The user's request concerns cash withdrawal charge.\", \"candidate_a45e27\": \"The user's request concerns unable to verify identity.\", \"candidate_a261d5\": \"The user's request concerns failed transfer.\", \"candidate_ac2188\": \"The user's request concerns top up limits.\", \"candidate_dbf8be\": \"The user's request concerns direct debit payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_132d93\": 0.0, \"candidate_748ae1\": 0.0, \"candidate_542ef7\": 0.0, \"candidate_a28814\": 1.0, \"candidate_a45e27\": 0.0, \"candidate_a261d5\": 0.0, \"candidate_ac2188\": 0.0, \"candidate_dbf8be\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6f9b49e908c739bfd3efcc7e:choice", "group_id": "fresh12:banking77:6f9b49e908c739bfd3efcc7e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why did I get a fee?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_53181a\": \"The user's request concerns pending transfer.\", \"candidate_fdce57\": \"The user's request concerns compromised card.\", \"candidate_4fa177\": \"The user's request concerns cash withdrawal charge.\", \"candidate_a2919a\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_e72558\": \"The user's request concerns supported cards and currencies.\", \"candidate_9b05ff\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_6695ec\": \"The user's request concerns top up limits.\", \"candidate_1b893a\": \"The user's request concerns activate my card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_53181a\": 0.0, \"candidate_fdce57\": 0.0, \"candidate_4fa177\": 1.0, \"candidate_a2919a\": 0.0, \"candidate_e72558\": 0.0, \"candidate_9b05ff\": 0.0, \"candidate_6695ec\": 0.0, \"candidate_1b893a\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c95e1bea56035e5dbcac9d6f:choice", "group_id": "fresh12:banking77:c95e1bea56035e5dbcac9d6f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Why is there a fee for withdrawing?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_357e48\": \"The user's request concerns apple pay or google pay.\", \"candidate_b24c5f\": \"The user's request concerns pending cash withdrawal.\", \"candidate_6c3caa\": \"The user's request concerns transfer timing.\", \"candidate_8aad79\": \"The user's request concerns card payment fee charged.\", \"candidate_90c61a\": \"The user's request concerns pending top up.\", \"candidate_3f711d\": \"The user's request concerns edit personal details.\", \"candidate_60f2e2\": \"The user's request concerns cash withdrawal charge.\", \"candidate_0bafed\": \"The user's request concerns topping up by card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_357e48\": 0.0, \"candidate_b24c5f\": 0.0, \"candidate_6c3caa\": 0.0, \"candidate_8aad79\": 0.0, \"candidate_90c61a\": 0.0, \"candidate_3f711d\": 0.0, \"candidate_60f2e2\": 1.0, \"candidate_0bafed\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7448f9593e41194a93ef58a3:choice", "group_id": "fresh12:banking77:7448f9593e41194a93ef58a3", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My account got charged for taking out cash, why did this happen?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7f6358\": \"The user's request concerns beneficiary not allowed.\", \"candidate_cf4510\": \"The user's request concerns order physical card.\", \"candidate_54eb02\": \"The user's request concerns verify source of funds.\", \"candidate_ebed5c\": \"The user's request concerns receiving money.\", \"candidate_e98042\": \"The user's request concerns cash withdrawal charge.\", \"candidate_6def8f\": \"The user's request concerns activate my card.\", \"candidate_94bf3f\": \"The user's request concerns supported cards and currencies.\", \"candidate_b3951e\": \"The user's request concerns direct debit payment not recognised.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7f6358\": 0.0, \"candidate_cf4510\": 0.0, \"candidate_54eb02\": 0.0, \"candidate_ebed5c\": 0.0, \"candidate_e98042\": 1.0, \"candidate_6def8f\": 0.0, \"candidate_94bf3f\": 0.0, \"candidate_b3951e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:7f120ec6fe98270e9217e582:choice", "group_id": "fresh12:banking77:7f120ec6fe98270e9217e582", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I took cash and got charged a fee\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_53d1f8\": \"The user's request concerns cash withdrawal charge.\", \"candidate_c0f80f\": \"The user's request concerns pending cash withdrawal.\", \"candidate_51a7af\": \"The user's request concerns pending top up.\", \"candidate_e0ac9a\": \"The user's request concerns age limit.\", \"candidate_7b6d0c\": \"The user's request concerns request refund.\", \"candidate_535fc1\": \"The user's request concerns verify source of funds.\", \"candidate_c6a547\": \"The user's request concerns topping up by card.\", \"candidate_2392ba\": \"The user's request concerns card about to expire.\"}}}}", "targets_json": "{\"decision\": {\"candidate_53d1f8\": 1.0, \"candidate_c0f80f\": 0.0, \"candidate_51a7af\": 0.0, \"candidate_e0ac9a\": 0.0, \"candidate_7b6d0c\": 0.0, \"candidate_535fc1\": 0.0, \"candidate_c6a547\": 0.0, \"candidate_2392ba\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:672ee5ccbc7989924a231049:choice", "group_id": "fresh12:banking77:672ee5ccbc7989924a231049", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My card is about to expire. How do I get a new one?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_90a39e\": \"The user's request concerns automatic top up.\", \"candidate_8d7540\": \"The user's request concerns transfer into account.\", \"candidate_2496b7\": \"The user's request concerns beneficiary not allowed.\", \"candidate_07e8b9\": \"The user's request concerns verify top up.\", \"candidate_b831e9\": \"The user's request concerns card about to expire.\", \"candidate_bba232\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_b05f7b\": \"The user's request concerns apple pay or google pay.\", \"candidate_678a40\": \"The user's request concerns cash withdrawal charge.\"}}}}", "targets_json": "{\"decision\": {\"candidate_90a39e\": 0.0, \"candidate_8d7540\": 0.0, \"candidate_2496b7\": 0.0, \"candidate_07e8b9\": 0.0, \"candidate_b831e9\": 1.0, \"candidate_bba232\": 0.0, \"candidate_b05f7b\": 0.0, \"candidate_678a40\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:52045b6e0acd334ebba18234:choice", "group_id": "fresh12:banking77:52045b6e0acd334ebba18234", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Since my card is about to expire, I need a new one.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_54e668\": \"The user's request concerns order physical card.\", \"candidate_b943ac\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_3f41b9\": \"The user's request concerns card arrival.\", \"candidate_eaa441\": \"The user's request concerns top up failed.\", \"candidate_f45428\": \"The user's request concerns get disposable virtual card.\", \"candidate_39ff8f\": \"The user's request concerns card about to expire.\", \"candidate_a23c4b\": \"The user's request concerns exchange rate.\", \"candidate_1c197f\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_54e668\": 0.0, \"candidate_b943ac\": 0.0, \"candidate_3f41b9\": 0.0, \"candidate_eaa441\": 0.0, \"candidate_f45428\": 0.0, \"candidate_39ff8f\": 1.0, \"candidate_a23c4b\": 0.0, \"candidate_1c197f\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:238259a34ef120e6dc28b22f:choice", "group_id": "fresh12:banking77:238259a34ef120e6dc28b22f", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"When my card expires what happens to my account?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_76d94d\": \"The user's request concerns disposable card limits.\", \"candidate_12bffb\": \"The user's request concerns exchange charge.\", \"candidate_033524\": \"The user's request concerns topping up by card.\", \"candidate_c76f85\": \"The user's request concerns country support.\", \"candidate_72502e\": \"The user's request concerns top up limits.\", \"candidate_a6c736\": \"The user's request concerns apple pay or google pay.\", \"candidate_8c57a8\": \"The user's request concerns order physical card.\", \"candidate_e4412f\": \"The user's request concerns card about to expire.\"}}}}", "targets_json": "{\"decision\": {\"candidate_76d94d\": 0.0, \"candidate_12bffb\": 0.0, \"candidate_033524\": 0.0, \"candidate_c76f85\": 0.0, \"candidate_72502e\": 0.0, \"candidate_a6c736\": 0.0, \"candidate_8c57a8\": 0.0, \"candidate_e4412f\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:427d8e58bb56f13419c1d8e1:choice", "group_id": "fresh12:banking77:427d8e58bb56f13419c1d8e1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need a new card since my old one is about to expire.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_9c7230\": \"The user's request concerns cancel transfer.\", \"candidate_9dc0b5\": \"The user's request concerns edit personal details.\", \"candidate_fda0d1\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_e8bf16\": \"The user's request concerns pending transfer.\", \"candidate_343fb0\": \"The user's request concerns card about to expire.\", \"candidate_186bee\": \"The user's request concerns transfer not received by recipient.\", \"candidate_bc143e\": \"The user's request concerns get disposable virtual card.\", \"candidate_284af4\": \"The user's request concerns transfer fee charged.\"}}}}", "targets_json": "{\"decision\": {\"candidate_9c7230\": 0.0, \"candidate_9dc0b5\": 0.0, \"candidate_fda0d1\": 0.0, \"candidate_e8bf16\": 0.0, \"candidate_343fb0\": 1.0, \"candidate_186bee\": 0.0, \"candidate_bc143e\": 0.0, \"candidate_284af4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:117d9e15d3e11bc3eee79ed6:choice", "group_id": "fresh12:banking77:117d9e15d3e11bc3eee79ed6", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can I receive a new card while I am in China?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_a32737\": \"The user's request concerns card payment not recognised.\", \"candidate_103293\": \"The user's request concerns getting spare card.\", \"candidate_4ede03\": \"The user's request concerns automatic top up.\", \"candidate_0ff2f0\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_2244fd\": \"The user's request concerns declined transfer.\", \"candidate_83e692\": \"The user's request concerns reverted card payment?.\", \"candidate_074a81\": \"The user's request concerns pending top up.\", \"candidate_490367\": \"The user's request concerns card about to expire.\"}}}}", "targets_json": "{\"decision\": {\"candidate_a32737\": 0.0, \"candidate_103293\": 0.0, \"candidate_4ede03\": 0.0, \"candidate_0ff2f0\": 0.0, \"candidate_2244fd\": 0.0, \"candidate_83e692\": 0.0, \"candidate_074a81\": 0.0, \"candidate_490367\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:0a0e5043a8803bf57b1182ca:choice", "group_id": "fresh12:banking77:0a0e5043a8803bf57b1182ca", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need a new card, my current card is nearly expired.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f8598a\": \"The user's request concerns card payment wrong exchange rate.\", \"candidate_6c6c70\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_ee742a\": \"The user's request concerns transfer not received by recipient.\", \"candidate_10ca12\": \"The user's request concerns card linking.\", \"candidate_4ced1b\": \"The user's request concerns top up by cash or cheque.\", \"candidate_41805f\": \"The user's request concerns fiat currency support.\", \"candidate_8c046e\": \"The user's request concerns top up by card charge.\", \"candidate_a228b9\": \"The user's request concerns card about to expire.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f8598a\": 0.0, \"candidate_6c6c70\": 0.0, \"candidate_ee742a\": 0.0, \"candidate_10ca12\": 0.0, \"candidate_4ced1b\": 0.0, \"candidate_41805f\": 0.0, \"candidate_8c046e\": 0.0, \"candidate_a228b9\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:01fdd0bba2984bd20e97a537:choice", "group_id": "fresh12:banking77:01fdd0bba2984bd20e97a537", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"can google pay be used to make a top-up\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_2fcb27\": \"The user's request concerns topping up by card.\", \"candidate_e661e3\": \"The user's request concerns transfer fee charged.\", \"candidate_0f4918\": \"The user's request concerns card linking.\", \"candidate_943813\": \"The user's request concerns card about to expire.\", \"candidate_12ae93\": \"The user's request concerns apple pay or google pay.\", \"candidate_882668\": \"The user's request concerns top up by cash or cheque.\", \"candidate_342af2\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_829b48\": \"The user's request concerns visa or mastercard.\"}}}}", "targets_json": "{\"decision\": {\"candidate_2fcb27\": 0.0, \"candidate_e661e3\": 0.0, \"candidate_0f4918\": 0.0, \"candidate_943813\": 0.0, \"candidate_12ae93\": 1.0, \"candidate_882668\": 0.0, \"candidate_342af2\": 0.0, \"candidate_829b48\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:3b8e56399f08b7b144f0250b:choice", "group_id": "fresh12:banking77:3b8e56399f08b7b144f0250b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"How do I top up? Can I use my apple watch?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_90a3e8\": \"The user's request concerns cancel transfer.\", \"candidate_4ca5dc\": \"The user's request concerns pin blocked.\", \"candidate_86814d\": \"The user's request concerns verify my identity.\", \"candidate_c4cf64\": \"The user's request concerns transfer not received by recipient.\", \"candidate_b6e9fb\": \"The user's request concerns apple pay or google pay.\", \"candidate_d53b3d\": \"The user's request concerns card payment fee charged.\", \"candidate_5517bd\": \"The user's request concerns supported cards and currencies.\", \"candidate_a29b0d\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_90a3e8\": 0.0, \"candidate_4ca5dc\": 0.0, \"candidate_86814d\": 0.0, \"candidate_c4cf64\": 0.0, \"candidate_b6e9fb\": 1.0, \"candidate_d53b3d\": 0.0, \"candidate_5517bd\": 0.0, \"candidate_a29b0d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:094ac0c1a5d39f0de7a622dc:choice", "group_id": "fresh12:banking77:094ac0c1a5d39f0de7a622dc", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is the top up feature available on the Apple Watch?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_898cdf\": \"The user's request concerns Refund not showing up.\", \"candidate_09ed5b\": \"The user's request concerns verify top up.\", \"candidate_b50a8b\": \"The user's request concerns supported cards and currencies.\", \"candidate_5e37e7\": \"The user's request concerns card delivery estimate.\", \"candidate_42f4a7\": \"The user's request concerns pending transfer.\", \"candidate_563063\": \"The user's request concerns top up limits.\", \"candidate_80cb37\": \"The user's request concerns apple pay or google pay.\", \"candidate_79dd77\": \"The user's request concerns unable to verify identity.\"}}}}", "targets_json": "{\"decision\": {\"candidate_898cdf\": 0.0, \"candidate_09ed5b\": 0.0, \"candidate_b50a8b\": 0.0, \"candidate_5e37e7\": 0.0, \"candidate_42f4a7\": 0.0, \"candidate_563063\": 0.0, \"candidate_80cb37\": 1.0, \"candidate_79dd77\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:ca679b143dd284328d8254c1:choice", "group_id": "fresh12:banking77:ca679b143dd284328d8254c1", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I got my American Express in Apple Bay but top up is not working\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_ebdbe3\": \"The user's request concerns declined cash withdrawal.\", \"candidate_430bc9\": \"The user's request concerns get physical card.\", \"candidate_5ac96b\": \"The user's request concerns getting virtual card.\", \"candidate_3ad453\": \"The user's request concerns compromised card.\", \"candidate_aeaad7\": \"The user's request concerns apple pay or google pay.\", \"candidate_4ad3bc\": \"The user's request concerns top up by cash or cheque.\", \"candidate_d1ee6c\": \"The user's request concerns pin blocked.\", \"candidate_cfd31b\": \"The user's request concerns automatic top up.\"}}}}", "targets_json": "{\"decision\": {\"candidate_ebdbe3\": 0.0, \"candidate_430bc9\": 0.0, \"candidate_5ac96b\": 0.0, \"candidate_3ad453\": 0.0, \"candidate_aeaad7\": 1.0, \"candidate_4ad3bc\": 0.0, \"candidate_d1ee6c\": 0.0, \"candidate_cfd31b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:57de6da9b9639b08a0b08ce8:choice", "group_id": "fresh12:banking77:57de6da9b9639b08a0b08ce8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"My American Express is in my Apple Pay and the top up is failing, why?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6177ab\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_d7f2d1\": \"The user's request concerns transfer fee charged.\", \"candidate_469b86\": \"The user's request concerns unable to verify identity.\", \"candidate_51072f\": \"The user's request concerns Refund not showing up.\", \"candidate_2b79f3\": \"The user's request concerns pending transfer.\", \"candidate_7c24f8\": \"The user's request concerns activate my card.\", \"candidate_0f9bf7\": \"The user's request concerns compromised card.\", \"candidate_cbac7f\": \"The user's request concerns apple pay or google pay.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6177ab\": 0.0, \"candidate_d7f2d1\": 0.0, \"candidate_469b86\": 0.0, \"candidate_51072f\": 0.0, \"candidate_2b79f3\": 0.0, \"candidate_7c24f8\": 0.0, \"candidate_0f9bf7\": 0.0, \"candidate_cbac7f\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1e086d65b1021acb01b63465:choice", "group_id": "fresh12:banking77:1e086d65b1021acb01b63465", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Does Google Play have an app to Top up?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_098ea4\": \"The user's request concerns atm support.\", \"candidate_8116f1\": \"The user's request concerns getting virtual card.\", \"candidate_f4ce4b\": \"The user's request concerns apple pay or google pay.\", \"candidate_646087\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_5ef2e5\": \"The user's request concerns age limit.\", \"candidate_489dad\": \"The user's request concerns card delivery estimate.\", \"candidate_13be72\": \"The user's request concerns declined card payment.\", \"candidate_c022a6\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_098ea4\": 0.0, \"candidate_8116f1\": 0.0, \"candidate_f4ce4b\": 1.0, \"candidate_646087\": 0.0, \"candidate_5ef2e5\": 0.0, \"candidate_489dad\": 0.0, \"candidate_13be72\": 0.0, \"candidate_c022a6\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:24ff0922a283458c0f92cfb8:choice", "group_id": "fresh12:banking77:24ff0922a283458c0f92cfb8", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What can I do to verify my identity?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_5add7c\": \"The user's request concerns top up failed.\", \"candidate_a22413\": \"The user's request concerns exchange charge.\", \"candidate_85647b\": \"The user's request concerns cancel transfer.\", \"candidate_bf25dd\": \"The user's request concerns verify my identity.\", \"candidate_67d86b\": \"The user's request concerns visa or mastercard.\", \"candidate_da00a8\": \"The user's request concerns unable to verify identity.\", \"candidate_f67872\": \"The user's request concerns pending card payment.\", \"candidate_b9e780\": \"The user's request concerns transfer not received by recipient.\"}}}}", "targets_json": "{\"decision\": {\"candidate_5add7c\": 0.0, \"candidate_a22413\": 0.0, \"candidate_85647b\": 0.0, \"candidate_bf25dd\": 1.0, \"candidate_67d86b\": 0.0, \"candidate_da00a8\": 0.0, \"candidate_f67872\": 0.0, \"candidate_b9e780\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:5337c9b454f1a20106d6f007:choice", "group_id": "fresh12:banking77:5337c9b454f1a20106d6f007", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What do I need to show who I am?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_4a2b96\": \"The user's request concerns topping up by card.\", \"candidate_15bd0f\": \"The user's request concerns compromised card.\", \"candidate_bbbabe\": \"The user's request concerns exchange charge.\", \"candidate_214553\": \"The user's request concerns verify my identity.\", \"candidate_a6642a\": \"The user's request concerns verify top up.\", \"candidate_a09b24\": \"The user's request concerns transfer not received by recipient.\", \"candidate_dc542d\": \"The user's request concerns disposable card limits.\", \"candidate_3bbba2\": \"The user's request concerns virtual card not working.\"}}}}", "targets_json": "{\"decision\": {\"candidate_4a2b96\": 0.0, \"candidate_15bd0f\": 0.0, \"candidate_bbbabe\": 0.0, \"candidate_214553\": 1.0, \"candidate_a6642a\": 0.0, \"candidate_a09b24\": 0.0, \"candidate_dc542d\": 0.0, \"candidate_3bbba2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:4d6b1e96cc5fe8c8723a3b71:choice", "group_id": "fresh12:banking77:4d6b1e96cc5fe8c8723a3b71", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What should I do to verify my identity?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_6c0428\": \"The user's request concerns verify source of funds.\", \"candidate_718e00\": \"The user's request concerns apple pay or google pay.\", \"candidate_713142\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_94fa02\": \"The user's request concerns pending top up.\", \"candidate_2c377b\": \"The user's request concerns verify my identity.\", \"candidate_c4fb3f\": \"The user's request concerns virtual card not working.\", \"candidate_2ed096\": \"The user's request concerns balance not updated after bank transfer.\", \"candidate_2361a2\": \"The user's request concerns balance not updated after cheque or cash deposit.\"}}}}", "targets_json": "{\"decision\": {\"candidate_6c0428\": 0.0, \"candidate_718e00\": 0.0, \"candidate_713142\": 0.0, \"candidate_94fa02\": 0.0, \"candidate_2c377b\": 1.0, \"candidate_c4fb3f\": 0.0, \"candidate_2ed096\": 0.0, \"candidate_2361a2\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:c5c50151e5ad723b072f0951:choice", "group_id": "fresh12:banking77:c5c50151e5ad723b072f0951", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"When getting my ID checked, what are the steps involved?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_c6b159\": \"The user's request concerns pin blocked.\", \"candidate_e4f461\": \"The user's request concerns verify my identity.\", \"candidate_c2b5a2\": \"The user's request concerns getting virtual card.\", \"candidate_865c5b\": \"The user's request concerns transfer not received by recipient.\", \"candidate_10bbda\": \"The user's request concerns balance not updated after cheque or cash deposit.\", \"candidate_1e9f85\": \"The user's request concerns transfer into account.\", \"candidate_a77f14\": \"The user's request concerns reverted card payment?.\", \"candidate_62f43b\": \"The user's request concerns card linking.\"}}}}", "targets_json": "{\"decision\": {\"candidate_c6b159\": 0.0, \"candidate_e4f461\": 1.0, \"candidate_c2b5a2\": 0.0, \"candidate_865c5b\": 0.0, \"candidate_10bbda\": 0.0, \"candidate_1e9f85\": 0.0, \"candidate_a77f14\": 0.0, \"candidate_62f43b\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:39ff2c95f01011110874bfa4:choice", "group_id": "fresh12:banking77:39ff2c95f01011110874bfa4", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Do I need any kind of documentation for the identity check?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_341a3d\": \"The user's request concerns activate my card.\", \"candidate_eed5b2\": \"The user's request concerns topping up by card.\", \"candidate_397259\": \"The user's request concerns automatic top up.\", \"candidate_c262e6\": \"The user's request concerns verify my identity.\", \"candidate_83ad3b\": \"The user's request concerns transfer timing.\", \"candidate_136010\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_ab7dfe\": \"The user's request concerns supported cards and currencies.\", \"candidate_5a0cee\": \"The user's request concerns getting virtual card.\"}}}}", "targets_json": "{\"decision\": {\"candidate_341a3d\": 0.0, \"candidate_eed5b2\": 0.0, \"candidate_397259\": 0.0, \"candidate_c262e6\": 1.0, \"candidate_83ad3b\": 0.0, \"candidate_136010\": 0.0, \"candidate_ab7dfe\": 0.0, \"candidate_5a0cee\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a03607d64b12cc1586a2df8e:choice", "group_id": "fresh12:banking77:a03607d64b12cc1586a2df8e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is there any documentation needed for the identity check?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_628fd3\": \"The user's request concerns transfer fee charged.\", \"candidate_a6b08b\": \"The user's request concerns card payment fee charged.\", \"candidate_da1d12\": \"The user's request concerns card about to expire.\", \"candidate_2a615a\": \"The user's request concerns declined cash withdrawal.\", \"candidate_ff72d7\": \"The user's request concerns Refund not showing up.\", \"candidate_54feed\": \"The user's request concerns exchange rate.\", \"candidate_884c49\": \"The user's request concerns verify my identity.\", \"candidate_eceaa5\": \"The user's request concerns pin blocked.\"}}}}", "targets_json": "{\"decision\": {\"candidate_628fd3\": 0.0, \"candidate_a6b08b\": 0.0, \"candidate_da1d12\": 0.0, \"candidate_2a615a\": 0.0, \"candidate_ff72d7\": 0.0, \"candidate_54feed\": 0.0, \"candidate_884c49\": 1.0, \"candidate_eceaa5\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:dec44987f28aa807c28fcc2e:choice", "group_id": "fresh12:banking77:dec44987f28aa807c28fcc2e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What locations are you in?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_060f18\": \"The user's request concerns pending transfer.\", \"candidate_a79728\": \"The user's request concerns top up by bank transfer charge.\", \"candidate_7c8ac6\": \"The user's request concerns card payment not recognised.\", \"candidate_081fb1\": \"The user's request concerns country support.\", \"candidate_b919c4\": \"The user's request concerns top up by cash or cheque.\", \"candidate_32c2ec\": \"The user's request concerns virtual card not working.\", \"candidate_1080a9\": \"The user's request concerns card arrival.\", \"candidate_16047d\": \"The user's request concerns visa or mastercard.\"}}}}", "targets_json": "{\"decision\": {\"candidate_060f18\": 0.0, \"candidate_a79728\": 0.0, \"candidate_7c8ac6\": 0.0, \"candidate_081fb1\": 1.0, \"candidate_b919c4\": 0.0, \"candidate_32c2ec\": 0.0, \"candidate_1080a9\": 0.0, \"candidate_16047d\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e85766b8bc490f61f0142229:choice", "group_id": "fresh12:banking77:e85766b8bc490f61f0142229", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What countries are supported?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_94e532\": \"The user's request concerns getting virtual card.\", \"candidate_c5a7c4\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_16626f\": \"The user's request concerns card arrival.\", \"candidate_ccb617\": \"The user's request concerns cancel transfer.\", \"candidate_985a4b\": \"The user's request concerns lost or stolen phone.\", \"candidate_f393e9\": \"The user's request concerns extra charge on statement.\", \"candidate_3a807f\": \"The user's request concerns country support.\", \"candidate_38f6f4\": \"The user's request concerns top up reverted.\"}}}}", "targets_json": "{\"decision\": {\"candidate_94e532\": 0.0, \"candidate_c5a7c4\": 0.0, \"candidate_16626f\": 0.0, \"candidate_ccb617\": 0.0, \"candidate_985a4b\": 0.0, \"candidate_f393e9\": 0.0, \"candidate_3a807f\": 1.0, \"candidate_38f6f4\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:1e795202260ae56cf11e142b:choice", "group_id": "fresh12:banking77:1e795202260ae56cf11e142b", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"What countries will my card be supported in?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_01d58f\": \"The user's request concerns card payment not recognised.\", \"candidate_a7a8ab\": \"The user's request concerns declined cash withdrawal.\", \"candidate_35bad4\": \"The user's request concerns contactless not working.\", \"candidate_6ae489\": \"The user's request concerns get disposable virtual card.\", \"candidate_c5142c\": \"The user's request concerns country support.\", \"candidate_21994b\": \"The user's request concerns get physical card.\", \"candidate_1b5d16\": \"The user's request concerns transfer fee charged.\", \"candidate_585db9\": \"The user's request concerns pending card payment.\"}}}}", "targets_json": "{\"decision\": {\"candidate_01d58f\": 0.0, \"candidate_a7a8ab\": 0.0, \"candidate_35bad4\": 0.0, \"candidate_6ae489\": 0.0, \"candidate_c5142c\": 1.0, \"candidate_21994b\": 0.0, \"candidate_1b5d16\": 0.0, \"candidate_585db9\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:6c339f7e0251698634325594:choice", "group_id": "fresh12:banking77:6c339f7e0251698634325594", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Can the card be mailed and used in Europe?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_f338bc\": \"The user's request concerns country support.\", \"candidate_7b70d3\": \"The user's request concerns topping up by card.\", \"candidate_13e8fc\": \"The user's request concerns verify top up.\", \"candidate_f43eee\": \"The user's request concerns transfer fee charged.\", \"candidate_47884e\": \"The user's request concerns transfer into account.\", \"candidate_f4aef0\": \"The user's request concerns direct debit payment not recognised.\", \"candidate_7edb50\": \"The user's request concerns top up by card charge.\", \"candidate_3ed658\": \"The user's request concerns failed transfer.\"}}}}", "targets_json": "{\"decision\": {\"candidate_f338bc\": 1.0, \"candidate_7b70d3\": 0.0, \"candidate_13e8fc\": 0.0, \"candidate_f43eee\": 0.0, \"candidate_47884e\": 0.0, \"candidate_f4aef0\": 0.0, \"candidate_7edb50\": 0.0, \"candidate_3ed658\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:e273d0ab0072de1c4a8d926e:choice", "group_id": "fresh12:banking77:e273d0ab0072de1c4a8d926e", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"Is it possible to get a card if I'm not in the UK?\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_7bfc9a\": \"The user's request concerns transfer into account.\", \"candidate_7ed9d2\": \"The user's request concerns activate my card.\", \"candidate_a4519b\": \"The user's request concerns contactless not working.\", \"candidate_3e79ec\": \"The user's request concerns cash withdrawal not recognised.\", \"candidate_4e08df\": \"The user's request concerns card swallowed.\", \"candidate_8d7eae\": \"The user's request concerns country support.\", \"candidate_afa1ba\": \"The user's request concerns exchange via app.\", \"candidate_c3e83e\": \"The user's request concerns transfer not received by recipient.\"}}}}", "targets_json": "{\"decision\": {\"candidate_7bfc9a\": 0.0, \"candidate_7ed9d2\": 0.0, \"candidate_a4519b\": 0.0, \"candidate_3e79ec\": 0.0, \"candidate_4e08df\": 0.0, \"candidate_8d7eae\": 1.0, \"candidate_afa1ba\": 0.0, \"candidate_c3e83e\": 0.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}
{"id": "fresh12:banking77:a498031a546d558c8f9ece08:choice", "group_id": "fresh12:banking77:a498031a546d558c8f9ece08", "split": "test", "family": "text_intent", "language": "en", "request_json": "{\"model\": \"Qwen/Qwen3.5-2B\", \"state\": {\"text\": \"I need a card, but I'm in the US at the moment.\", \"images\": []}, \"questions\": {\"decision\": {\"type\": \"choice\", \"instructions\": \"Which outcome follows from the evidence and any supplied rules? Return probabilities when facts are uncertain.\", \"criteria\": {\"candidate_975579\": \"The user's request concerns exchange charge.\", \"candidate_1621c6\": \"The user's request concerns change pin.\", \"candidate_cd4b29\": \"The user's request concerns visa or mastercard.\", \"candidate_844c2c\": \"The user's request concerns beneficiary not allowed.\", \"candidate_4eaf01\": \"The user's request concerns verify my identity.\", \"candidate_6c0bfb\": \"The user's request concerns activate my card.\", \"candidate_83b71d\": \"The user's request concerns top up failed.\", \"candidate_e6e75d\": \"The user's request concerns country support.\"}}}}", "targets_json": "{\"decision\": {\"candidate_975579\": 0.0, \"candidate_1621c6\": 0.0, \"candidate_cd4b29\": 0.0, \"candidate_844c2c\": 0.0, \"candidate_4eaf01\": 0.0, \"candidate_6c0bfb\": 0.0, \"candidate_83b71d\": 0.0, \"candidate_e6e75d\": 1.0}}", "source_id": "PolyAI/banking77", "source_revision": "57ec275d8078af65b7731c2a98be812d844a6d6b", "source_license": "CC-BY-4.0", "source_url": "https://huggingface.co/datasets/PolyAI/banking77", "image_sha256": null, "image_paths_json": "[]"}