Text Classification
Transformers
ONNX
Safetensors
modernbert
prompt-injection
jailbreak-detection
llm-security
ai-agents
guardrails
onnxruntime
mcp
text-embeddings-inference
Instructions to use usezn/Galvanize-2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use usezn/Galvanize-2 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="usezn/Galvanize-2")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("usezn/Galvanize-2") model = AutoModelForSequenceClassification.from_pretrained("usezn/Galvanize-2", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download benchmark/results.json from usezn/Galvanize-2: direct link, hf CLI and curl.
- Browser
- Download file 224 kB
-
https://huggingface.co/usezn/Galvanize-2/resolve/main/benchmark/results.json
- Command line
-
hf download hf://usezn/Galvanize-2/benchmark/results.json
-
curl -L -o results.json https://huggingface.co/usezn/Galvanize-2/resolve/main/benchmark/results.json
224 kB
| { | |
| "benchmark": "Galvanize 2 vs open prompt-injection detectors (public eval suite)", | |
| "run": { | |
| "date_utc": "2026-10-12T05:29:16Z", | |
| "hardware": "1x NVIDIA L4 GPU, fp32, TF32 off", | |
| "device": "cuda", | |
| "torch": "2.11.0+cu130", | |
| "transformers": "5.18.0", | |
| "python": "3.13.15", | |
| "protocols": [ | |
| "serve", | |
| "hz" | |
| ], | |
| "limit": null | |
| }, | |
| "suite": { | |
| "repo": "Horizon-Labs/prompt-injection-eval-suite", | |
| "revision": "27c7bcd6f46e842f4c4d8b114b30eb6b1d892965", | |
| "sha256": { | |
| "notinject": "75d8a7bfde6172335d6178dbfba192209570309a19edfdb11d2120a3b6e919d0", | |
| "xstest": "3ca278fe64310fa5172d34cd17eef9ef8cbe2d20300756c6424d07d41154b90c", | |
| "orbench_hard": "ceaa7b010b08377187e637d2f7c61912914fa0b8f3788cabdcfbc1c16b3c85a8", | |
| "deepset_test": "e2c1ed4700967312e2e95e87ae48c0e0617d02b0fed4335b0d9a631b66bb47d6", | |
| "jackhhao_test": "5a1987583bf94e528f8cfc98f21664eb1206af605f44979317e30a8c4a2b5a87", | |
| "simsonsun_jailbreaks": "967cdd37e16e5377e476706e4047d2871c930ca48dc829b4001abe76ec72a0dc", | |
| "boundary_pairs_test": "44e5d543bd4bef91249b91a5d996086631e4cae4d499616bed4202fd4b4b774d", | |
| "bipia": "744ab24ff443589df2d7ea22fc42618f124f5682487938021f35bdf08c8374bd", | |
| "piarena": "0bbd59a14b622e0ed8630cac365c136537c2ab2b31e4b3fd9587b5edc0b88b9a", | |
| "llmail_phase2": "707adf5b915a7f59362a50ccce0652f9b48a145c3a8927762e06c5a777552a9f", | |
| "neuralchemy_test": "b8228d74858b2e8d2e4e3c9589423e9758fc4b17e1e5a67a00b4bce812a8bcb9", | |
| "slabs_test": "67fab0e0cf485f6530f53fdf291a89f27c75e01f674e78ae1a61286642b02f90", | |
| "agentic5k_test": "a1adbc4052b57594be26bea9a4064d91bc2291297cf452760ea9938fb7aab14a" | |
| }, | |
| "card_metric": { | |
| "notinject": "acc", | |
| "xstest": "acc", | |
| "orbench_hard": "acc", | |
| "jackhhao_test": "f1", | |
| "deepset_test": "f1", | |
| "simsonsun_jailbreaks": "recall", | |
| "boundary_pairs_test": "f1", | |
| "bipia": "f1", | |
| "piarena": "f1", | |
| "llmail_phase2": "recall" | |
| }, | |
| "supplementary_metric": { | |
| "neuralchemy_test": "f1", | |
| "slabs_test": "f1", | |
| "agentic5k_test": "f1" | |
| } | |
| }, | |
| "protocols": { | |
| "serve": "re.sub(r\"\\s+\", \" \", text).strip(), model tokenizer, first 512 tokens, one window, fp32", | |
| "hz": "Horizon's evaluate.py: raw text, window min(2048, max positions) (512 for DeBERTa and Galvanize-60M), stride window // 4, document score = max over windows, fp32" | |
| }, | |
| "subsets": { | |
| "decontaminated": "rows listed in decontam_excluded_ids.json removed", | |
| "full": "every row" | |
| }, | |
| "models": { | |
| "galvanize-2": { | |
| "name": "Galvanize 2", | |
| "repo": "usezn/Galvanize-2", | |
| "revision": null, | |
| "weights_sha256": "3f474d28f4d39ca3ba95fba760fd5daa4687952de89ee5c1232f66cd99357eee", | |
| "threshold": 0.661988004118821, | |
| "positive_label": "INJECTION", | |
| "model_type": "modernbert", | |
| "hz_window": 2048, | |
| "error": null, | |
| "note": "scored from the model.safetensors file with this SHA-256; run_benchmark.py loads usezn/Galvanize-2 and refuses any other file" | |
| }, | |
| "horizon-small": { | |
| "name": "Horizon small (141M)", | |
| "repo": "Horizon-Labs/prompt-injection-guard-small", | |
| "revision": "e0e88c4e940ee7e1b52bdf943d501ec696b83b9b", | |
| "weights_sha256": "a23196c117baf414b4b8abd138c5f5656e728877c0bdb2775c361ee89df16deb", | |
| "threshold": 0.5, | |
| "positive_label": "INJECTION", | |
| "model_type": "modernbert", | |
| "hz_window": 2048, | |
| "error": null | |
| }, | |
| "horizon-base": { | |
| "name": "Horizon base (308M)", | |
| "repo": "Horizon-Labs/prompt-injection-guard-base", | |
| "revision": "e4fbfc6f603e24c675cdf310c3b8d7abaa825666", | |
| "weights_sha256": "d0c0f899f118481855cfe34ba718cf5e44613d266270721c380d54741626d405", | |
| "threshold": 0.5, | |
| "positive_label": "INJECTION", | |
| "model_type": "modernbert", | |
| "hz_window": 2048, | |
| "error": null | |
| }, | |
| "horizon-large": { | |
| "name": "Horizon large", | |
| "repo": "Horizon-Labs/prompt-injection-guard-large", | |
| "revision": "bbc67bafa71c791ad802d4df0ae7639d4e880dea", | |
| "weights_sha256": "81bd637f0964fea52d452ddda3c814cf7d822ca7db6b2a34015db668f4a183bf", | |
| "threshold": 0.5, | |
| "positive_label": "INJECTION", | |
| "model_type": "xlm-roberta", | |
| "hz_window": 2048, | |
| "error": null | |
| }, | |
| "galvanize-60m": { | |
| "name": "Galvanize-60M", | |
| "repo": "usezn/Galvanize-60M", | |
| "revision": "7bf8df0d3c9cd724c5773a1c73d7c81ebb0f2082", | |
| "weights_sha256": "53684cff19857acaa0fd8bb81c9271224d56e5ef1f9b4942a9ddaa846f1e4027", | |
| "threshold": 0.95, | |
| "positive_label": "MALICIOUS", | |
| "model_type": "galvanize-60m", | |
| "hz_window": 512, | |
| "error": null | |
| }, | |
| "protectai-v2": { | |
| "name": "ProtectAI v2", | |
| "repo": "protectai/deberta-v3-base-prompt-injection-v2", | |
| "revision": "90c9989b1a342275dd0d1a95aad283c04e075671", | |
| "weights_sha256": "6521cb8d0ac08148c81464899c424e6148fcc62befa371089fa4061d8b6e0424", | |
| "threshold": 0.5, | |
| "positive_label": "INJECTION", | |
| "model_type": "deberta-v2", | |
| "hz_window": 512, | |
| "error": null | |
| }, | |
| "deepset": { | |
| "name": "deepset", | |
| "repo": "deepset/deberta-v3-base-injection", | |
| "revision": "80dda00d0b0d9a03917a7685e2ddbcd28e04dbb1", | |
| "weights_sha256": "16862b96671d118a79b8d3a1352a552913797f9fc3cc7db2e06a05002abfc084", | |
| "threshold": 0.5, | |
| "positive_label": "INJECTION", | |
| "model_type": "deberta-v2", | |
| "hz_window": 512, | |
| "error": null | |
| }, | |
| "wolf-defender": { | |
| "name": "Wolf Defender", | |
| "repo": "patronus-studio/wolf-defender-prompt-injection", | |
| "revision": "a77e7b669f079c3ad7e72e4ad10f3bd173ac4f4d", | |
| "weights_sha256": "e8330cc19b083deb45847d873a69bc840cd8913b62a631289e2c034fa5b6cf3b", | |
| "threshold": 0.5, | |
| "positive_label": "injection", | |
| "model_type": "modernbert", | |
| "hz_window": 2048, | |
| "error": null | |
| }, | |
| "neuraltrust-small": { | |
| "name": "NeuralTrust small", | |
| "repo": "NeuralTrust/prompt-guard-oss-small", | |
| "revision": "40e5c56b68a1b081484c3e56c9bee726e2749138", | |
| "weights_sha256": "5b1ba9c313447d445146016125ab337a2b5298b2dc2198fcd20e20727a32382f", | |
| "threshold": 0.5, | |
| "positive_label": "jailbreak", | |
| "model_type": "modernbert", | |
| "hz_window": 2048, | |
| "error": null | |
| } | |
| }, | |
| "decontamination": { | |
| "notinject": { | |
| "rows": 339, | |
| "excluded": 0, | |
| "kept": 339, | |
| "training_overlap_exact_pct": 0.0, | |
| "training_overlap_normalised_pct": 0.0, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 0.0, | |
| "upstream_train_split_overlap_any_pct": 0.0 | |
| }, | |
| "xstest": { | |
| "rows": 450, | |
| "excluded": 9, | |
| "kept": 441, | |
| "training_overlap_exact_pct": 2.0, | |
| "training_overlap_normalised_pct": 2.0, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 2.0, | |
| "upstream_train_split_overlap_any_pct": 15.33 | |
| }, | |
| "orbench_hard": { | |
| "rows": 1319, | |
| "excluded": 7, | |
| "kept": 1312, | |
| "training_overlap_exact_pct": 0.53, | |
| "training_overlap_normalised_pct": 0.53, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 0.53, | |
| "upstream_train_split_overlap_any_pct": 99.92 | |
| }, | |
| "jackhhao_test": { | |
| "rows": 262, | |
| "excluded": 126, | |
| "kept": 136, | |
| "training_overlap_exact_pct": 16.03, | |
| "training_overlap_normalised_pct": 33.97, | |
| "training_overlap_head200_pct": 42.75, | |
| "training_overlap_any_pct": 48.09, | |
| "upstream_train_split_overlap_any_pct": 87.02 | |
| }, | |
| "deepset_test": { | |
| "rows": 116, | |
| "excluded": 0, | |
| "kept": 116, | |
| "training_overlap_exact_pct": 0.0, | |
| "training_overlap_normalised_pct": 0.0, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 0.0, | |
| "upstream_train_split_overlap_any_pct": 0.0 | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "rows": 2426, | |
| "excluded": 131, | |
| "kept": 2295, | |
| "training_overlap_exact_pct": 0.99, | |
| "training_overlap_normalised_pct": 2.68, | |
| "training_overlap_head200_pct": 5.07, | |
| "training_overlap_any_pct": 5.4, | |
| "upstream_train_split_overlap_any_pct": 4.95 | |
| }, | |
| "boundary_pairs_test": { | |
| "rows": 240, | |
| "excluded": 0, | |
| "kept": 240, | |
| "training_overlap_exact_pct": 0.0, | |
| "training_overlap_normalised_pct": 0.0, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 0.0, | |
| "upstream_train_split_overlap_any_pct": 0.0 | |
| }, | |
| "bipia": { | |
| "rows": 3178, | |
| "excluded": 0, | |
| "kept": 3178, | |
| "training_overlap_exact_pct": 0.0, | |
| "training_overlap_normalised_pct": 0.0, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 0.0, | |
| "upstream_train_split_overlap_any_pct": 0.0 | |
| }, | |
| "piarena": { | |
| "rows": 1500, | |
| "excluded": 0, | |
| "kept": 1500, | |
| "training_overlap_exact_pct": 0.0, | |
| "training_overlap_normalised_pct": 0.0, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 0.0, | |
| "upstream_train_split_overlap_any_pct": 0.0 | |
| }, | |
| "llmail_phase2": { | |
| "rows": 2000, | |
| "excluded": 0, | |
| "kept": 2000, | |
| "training_overlap_exact_pct": 0.0, | |
| "training_overlap_normalised_pct": 0.0, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 0.0, | |
| "upstream_train_split_overlap_any_pct": 0.0 | |
| }, | |
| "neuralchemy_test": { | |
| "rows": 942, | |
| "excluded": 2, | |
| "kept": 940, | |
| "training_overlap_exact_pct": 0.0, | |
| "training_overlap_normalised_pct": 0.0, | |
| "training_overlap_head200_pct": 0.21, | |
| "training_overlap_any_pct": 0.21, | |
| "upstream_train_split_overlap_any_pct": 1.06 | |
| }, | |
| "slabs_test": { | |
| "rows": 2101, | |
| "excluded": 0, | |
| "kept": 2101, | |
| "training_overlap_exact_pct": 0.0, | |
| "training_overlap_normalised_pct": 0.0, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 0.0, | |
| "upstream_train_split_overlap_any_pct": 0.0 | |
| }, | |
| "agentic5k_test": { | |
| "rows": 545, | |
| "excluded": 0, | |
| "kept": 545, | |
| "training_overlap_exact_pct": 0.0, | |
| "training_overlap_normalised_pct": 0.0, | |
| "training_overlap_head200_pct": 0.0, | |
| "training_overlap_any_pct": 0.0, | |
| "upstream_train_split_overlap_any_pct": 0.0 | |
| } | |
| }, | |
| "results": { | |
| "galvanize-2": { | |
| "serve": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 0.9823, | |
| "fpr": 0.0177 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9558, | |
| "fpr": 0.0442 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 0.9823, | |
| "fpr": 0.0177 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9558, | |
| "fpr": 0.0442 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 0.9985, | |
| "fpr": 0.0015 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9954, | |
| "fpr": 0.0046 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 0.9985, | |
| "fpr": 0.0015 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9955, | |
| "fpr": 0.0045 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.662, | |
| "acc": 0.9485, | |
| "recall": 0.8684, | |
| "fpr": 0.0204, | |
| "auroc": 0.9815, | |
| "pauc10": 0.9098, | |
| "precision": 0.9429, | |
| "f1": 0.9041 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9559, | |
| "recall": 0.8947, | |
| "fpr": 0.0204, | |
| "auroc": 0.9815, | |
| "pauc10": 0.9098, | |
| "precision": 0.9444, | |
| "f1": 0.9189 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.662, | |
| "acc": 0.9198, | |
| "recall": 0.8705, | |
| "fpr": 0.0244, | |
| "auroc": 0.9736, | |
| "pauc10": 0.8397, | |
| "precision": 0.9758, | |
| "f1": 0.9202 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9275, | |
| "recall": 0.8921, | |
| "fpr": 0.0325, | |
| "auroc": 0.9736, | |
| "pauc10": 0.8397, | |
| "precision": 0.9688, | |
| "f1": 0.9288 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.662, | |
| "acc": 0.7759, | |
| "recall": 0.5667, | |
| "fpr": 0.0, | |
| "auroc": 0.978, | |
| "pauc10": 0.9018, | |
| "precision": 1.0, | |
| "f1": 0.7234 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.978, | |
| "pauc10": 0.9018, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.662, | |
| "acc": 0.7759, | |
| "recall": 0.5667, | |
| "fpr": 0.0, | |
| "auroc": 0.978, | |
| "pauc10": 0.9018, | |
| "precision": 1.0, | |
| "f1": 0.7234 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.978, | |
| "pauc10": 0.9018, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.662, | |
| "acc": 0.7412, | |
| "recall": 0.7412 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.8174, | |
| "recall": 0.8174 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.662, | |
| "acc": 0.7477, | |
| "recall": 0.7477 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8232, | |
| "recall": 0.8232 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.662, | |
| "acc": 0.9833, | |
| "recall": 0.9917, | |
| "fpr": 0.025, | |
| "auroc": 0.993, | |
| "pauc10": 0.9299, | |
| "precision": 0.9754, | |
| "f1": 0.9835 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9792, | |
| "recall": 1.0, | |
| "fpr": 0.0417, | |
| "auroc": 0.993, | |
| "pauc10": 0.9299, | |
| "precision": 0.96, | |
| "f1": 0.9796 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.662, | |
| "acc": 0.9833, | |
| "recall": 0.9917, | |
| "fpr": 0.025, | |
| "auroc": 0.993, | |
| "pauc10": 0.9299, | |
| "precision": 0.9754, | |
| "f1": 0.9835 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9792, | |
| "recall": 1.0, | |
| "fpr": 0.0417, | |
| "auroc": 0.993, | |
| "pauc10": 0.9299, | |
| "precision": 0.96, | |
| "f1": 0.9796 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.662, | |
| "acc": 0.46, | |
| "recall": 0.4307, | |
| "fpr": 0.0449, | |
| "auroc": 0.8307, | |
| "pauc10": 0.4421, | |
| "precision": 0.9938, | |
| "f1": 0.6009 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5296, | |
| "recall": 0.5053, | |
| "fpr": 0.0618, | |
| "auroc": 0.8307, | |
| "pauc10": 0.4421, | |
| "precision": 0.9928, | |
| "f1": 0.6698 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.662, | |
| "acc": 0.46, | |
| "recall": 0.4307, | |
| "fpr": 0.0449, | |
| "auroc": 0.8307, | |
| "pauc10": 0.4421, | |
| "precision": 0.9938, | |
| "f1": 0.6009 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5296, | |
| "recall": 0.5053, | |
| "fpr": 0.0618, | |
| "auroc": 0.8307, | |
| "pauc10": 0.4421, | |
| "precision": 0.9928, | |
| "f1": 0.6698 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.662, | |
| "acc": 0.8453, | |
| "recall": 0.692, | |
| "fpr": 0.0013, | |
| "auroc": 0.9301, | |
| "pauc10": 0.7904, | |
| "precision": 0.9981, | |
| "f1": 0.8173 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8573, | |
| "recall": 0.716, | |
| "fpr": 0.0013, | |
| "auroc": 0.9301, | |
| "pauc10": 0.7904, | |
| "precision": 0.9981, | |
| "f1": 0.8339 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.662, | |
| "acc": 0.8453, | |
| "recall": 0.692, | |
| "fpr": 0.0013, | |
| "auroc": 0.9301, | |
| "pauc10": 0.7904, | |
| "precision": 0.9981, | |
| "f1": 0.8173 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8573, | |
| "recall": 0.716, | |
| "fpr": 0.0013, | |
| "auroc": 0.9301, | |
| "pauc10": 0.7904, | |
| "precision": 0.9981, | |
| "f1": 0.8339 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.662, | |
| "acc": 0.9435, | |
| "recall": 0.9435 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9535, | |
| "recall": 0.9535 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.662, | |
| "acc": 0.9435, | |
| "recall": 0.9435 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9535, | |
| "recall": 0.9535 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.662, | |
| "acc": 0.9479, | |
| "recall": 0.9309, | |
| "fpr": 0.0282, | |
| "auroc": 0.9915, | |
| "pauc10": 0.9489, | |
| "precision": 0.979, | |
| "f1": 0.9543 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9532, | |
| "recall": 0.9418, | |
| "fpr": 0.0308, | |
| "auroc": 0.9915, | |
| "pauc10": 0.9489, | |
| "precision": 0.9774, | |
| "f1": 0.9593 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.662, | |
| "acc": 0.948, | |
| "recall": 0.9312, | |
| "fpr": 0.0282, | |
| "auroc": 0.9915, | |
| "pauc10": 0.9488, | |
| "precision": 0.979, | |
| "f1": 0.9545 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9533, | |
| "recall": 0.942, | |
| "fpr": 0.0308, | |
| "auroc": 0.9915, | |
| "pauc10": 0.9488, | |
| "precision": 0.9774, | |
| "f1": 0.9594 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.662, | |
| "acc": 0.9248, | |
| "recall": 0.8506, | |
| "fpr": 0.001, | |
| "auroc": 0.9972, | |
| "pauc10": 0.9749, | |
| "precision": 0.9989, | |
| "f1": 0.9188 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9467, | |
| "recall": 0.8944, | |
| "fpr": 0.001, | |
| "auroc": 0.9972, | |
| "pauc10": 0.9749, | |
| "precision": 0.9989, | |
| "f1": 0.9438 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.662, | |
| "acc": 0.9248, | |
| "recall": 0.8506, | |
| "fpr": 0.001, | |
| "auroc": 0.9972, | |
| "pauc10": 0.9749, | |
| "precision": 0.9989, | |
| "f1": 0.9188 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9467, | |
| "recall": 0.8944, | |
| "fpr": 0.001, | |
| "auroc": 0.9972, | |
| "pauc10": 0.9749, | |
| "precision": 0.9989, | |
| "f1": 0.9438 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.662, | |
| "acc": 0.8422, | |
| "recall": 1.0, | |
| "fpr": 0.3245, | |
| "auroc": 0.9944, | |
| "pauc10": 0.9704, | |
| "precision": 0.765, | |
| "f1": 0.8669 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8202, | |
| "recall": 1.0, | |
| "fpr": 0.3698, | |
| "auroc": 0.9944, | |
| "pauc10": 0.9704, | |
| "precision": 0.7407, | |
| "f1": 0.8511 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.662, | |
| "acc": 0.8422, | |
| "recall": 1.0, | |
| "fpr": 0.3245, | |
| "auroc": 0.9944, | |
| "pauc10": 0.9704, | |
| "precision": 0.765, | |
| "f1": 0.8669 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8202, | |
| "recall": 1.0, | |
| "fpr": 0.3698, | |
| "auroc": 0.9944, | |
| "pauc10": 0.9704, | |
| "precision": 0.7407, | |
| "f1": 0.8511 | |
| } | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 0.9823, | |
| "fpr": 0.0177 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9558, | |
| "fpr": 0.0442 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 0.9823, | |
| "fpr": 0.0177 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9558, | |
| "fpr": 0.0442 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 0.9985, | |
| "fpr": 0.0015 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9954, | |
| "fpr": 0.0046 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.662, | |
| "acc": 0.9985, | |
| "fpr": 0.0015 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9955, | |
| "fpr": 0.0045 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.662, | |
| "acc": 0.9632, | |
| "recall": 0.8947, | |
| "fpr": 0.0102, | |
| "auroc": 0.9885, | |
| "pauc10": 0.9307, | |
| "precision": 0.9714, | |
| "f1": 0.9315 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9632, | |
| "recall": 0.9211, | |
| "fpr": 0.0204, | |
| "auroc": 0.9885, | |
| "pauc10": 0.9307, | |
| "precision": 0.9459, | |
| "f1": 0.9333 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.662, | |
| "acc": 0.9198, | |
| "recall": 0.8633, | |
| "fpr": 0.0163, | |
| "auroc": 0.9769, | |
| "pauc10": 0.8455, | |
| "precision": 0.9836, | |
| "f1": 0.9195 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9313, | |
| "recall": 0.8993, | |
| "fpr": 0.0325, | |
| "auroc": 0.9769, | |
| "pauc10": 0.8455, | |
| "precision": 0.969, | |
| "f1": 0.9328 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.662, | |
| "acc": 0.7672, | |
| "recall": 0.55, | |
| "fpr": 0.0, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8869, | |
| "precision": 1.0, | |
| "f1": 0.7097 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.819, | |
| "recall": 0.65, | |
| "fpr": 0.0, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8869, | |
| "precision": 1.0, | |
| "f1": 0.7879 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.662, | |
| "acc": 0.7672, | |
| "recall": 0.55, | |
| "fpr": 0.0, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8869, | |
| "precision": 1.0, | |
| "f1": 0.7097 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.819, | |
| "recall": 0.65, | |
| "fpr": 0.0, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8869, | |
| "precision": 1.0, | |
| "f1": 0.7879 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.662, | |
| "acc": 0.7028, | |
| "recall": 0.7028 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7821, | |
| "recall": 0.7821 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.662, | |
| "acc": 0.7098, | |
| "recall": 0.7098 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7885, | |
| "recall": 0.7885 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.662, | |
| "acc": 0.9833, | |
| "recall": 0.9917, | |
| "fpr": 0.025, | |
| "auroc": 0.993, | |
| "pauc10": 0.9299, | |
| "precision": 0.9754, | |
| "f1": 0.9835 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9792, | |
| "recall": 1.0, | |
| "fpr": 0.0417, | |
| "auroc": 0.993, | |
| "pauc10": 0.9299, | |
| "precision": 0.96, | |
| "f1": 0.9796 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.662, | |
| "acc": 0.9833, | |
| "recall": 0.9917, | |
| "fpr": 0.025, | |
| "auroc": 0.993, | |
| "pauc10": 0.9299, | |
| "precision": 0.9754, | |
| "f1": 0.9835 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9792, | |
| "recall": 1.0, | |
| "fpr": 0.0417, | |
| "auroc": 0.993, | |
| "pauc10": 0.9299, | |
| "precision": 0.96, | |
| "f1": 0.9796 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.662, | |
| "acc": 0.3493, | |
| "recall": 0.3117, | |
| "fpr": 0.0169, | |
| "auroc": 0.855, | |
| "pauc10": 0.4253, | |
| "precision": 0.9968, | |
| "f1": 0.4749 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.4198, | |
| "recall": 0.387, | |
| "fpr": 0.0281, | |
| "auroc": 0.855, | |
| "pauc10": 0.4253, | |
| "precision": 0.9957, | |
| "f1": 0.5574 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.662, | |
| "acc": 0.3493, | |
| "recall": 0.3117, | |
| "fpr": 0.0169, | |
| "auroc": 0.855, | |
| "pauc10": 0.4253, | |
| "precision": 0.9968, | |
| "f1": 0.4749 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.4198, | |
| "recall": 0.387, | |
| "fpr": 0.0281, | |
| "auroc": 0.855, | |
| "pauc10": 0.4253, | |
| "precision": 0.9957, | |
| "f1": 0.5574 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.662, | |
| "acc": 0.9287, | |
| "recall": 0.8573, | |
| "fpr": 0.0, | |
| "auroc": 0.9964, | |
| "pauc10": 0.9809, | |
| "precision": 1.0, | |
| "f1": 0.9232 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.942, | |
| "recall": 0.8853, | |
| "fpr": 0.0013, | |
| "auroc": 0.9964, | |
| "pauc10": 0.9809, | |
| "precision": 0.9985, | |
| "f1": 0.9385 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.662, | |
| "acc": 0.9287, | |
| "recall": 0.8573, | |
| "fpr": 0.0, | |
| "auroc": 0.9964, | |
| "pauc10": 0.9809, | |
| "precision": 1.0, | |
| "f1": 0.9232 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.942, | |
| "recall": 0.8853, | |
| "fpr": 0.0013, | |
| "auroc": 0.9964, | |
| "pauc10": 0.9809, | |
| "precision": 0.9985, | |
| "f1": 0.9385 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.662, | |
| "acc": 0.9815, | |
| "recall": 0.9815 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.988, | |
| "recall": 0.988 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.662, | |
| "acc": 0.9815, | |
| "recall": 0.9815 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.988, | |
| "recall": 0.988 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.662, | |
| "acc": 0.9479, | |
| "recall": 0.9309, | |
| "fpr": 0.0282, | |
| "auroc": 0.9914, | |
| "pauc10": 0.948, | |
| "precision": 0.979, | |
| "f1": 0.9543 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9521, | |
| "recall": 0.94, | |
| "fpr": 0.0308, | |
| "auroc": 0.9914, | |
| "pauc10": 0.948, | |
| "precision": 0.9773, | |
| "f1": 0.9583 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.662, | |
| "acc": 0.948, | |
| "recall": 0.9312, | |
| "fpr": 0.0282, | |
| "auroc": 0.9915, | |
| "pauc10": 0.948, | |
| "precision": 0.979, | |
| "f1": 0.9545 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9522, | |
| "recall": 0.9402, | |
| "fpr": 0.0308, | |
| "auroc": 0.9915, | |
| "pauc10": 0.948, | |
| "precision": 0.9774, | |
| "f1": 0.9584 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.662, | |
| "acc": 0.9248, | |
| "recall": 0.8506, | |
| "fpr": 0.001, | |
| "auroc": 0.9972, | |
| "pauc10": 0.9749, | |
| "precision": 0.9989, | |
| "f1": 0.9188 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9467, | |
| "recall": 0.8944, | |
| "fpr": 0.001, | |
| "auroc": 0.9972, | |
| "pauc10": 0.9749, | |
| "precision": 0.9989, | |
| "f1": 0.9438 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.662, | |
| "acc": 0.9248, | |
| "recall": 0.8506, | |
| "fpr": 0.001, | |
| "auroc": 0.9972, | |
| "pauc10": 0.9749, | |
| "precision": 0.9989, | |
| "f1": 0.9188 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9467, | |
| "recall": 0.8944, | |
| "fpr": 0.001, | |
| "auroc": 0.9972, | |
| "pauc10": 0.9749, | |
| "precision": 0.9989, | |
| "f1": 0.9438 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.662, | |
| "acc": 0.8422, | |
| "recall": 1.0, | |
| "fpr": 0.3245, | |
| "auroc": 0.9944, | |
| "pauc10": 0.9704, | |
| "precision": 0.765, | |
| "f1": 0.8669 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8202, | |
| "recall": 1.0, | |
| "fpr": 0.3698, | |
| "auroc": 0.9944, | |
| "pauc10": 0.9704, | |
| "precision": 0.7407, | |
| "f1": 0.8511 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.662, | |
| "acc": 0.8422, | |
| "recall": 1.0, | |
| "fpr": 0.3245, | |
| "auroc": 0.9944, | |
| "pauc10": 0.9704, | |
| "precision": 0.765, | |
| "f1": 0.8669 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8202, | |
| "recall": 1.0, | |
| "fpr": 0.3698, | |
| "auroc": 0.9944, | |
| "pauc10": 0.9704, | |
| "precision": 0.7407, | |
| "f1": 0.8511 | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "horizon-small": { | |
| "serve": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9617, | |
| "fpr": 0.0383 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9617, | |
| "fpr": 0.0383 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9617, | |
| "fpr": 0.0383 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9617, | |
| "fpr": 0.0383 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9947, | |
| "fpr": 0.0053 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9947, | |
| "fpr": 0.0053 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9947, | |
| "fpr": 0.0053 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9947, | |
| "fpr": 0.0053 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9338, | |
| "recall": 0.8947, | |
| "fpr": 0.051, | |
| "auroc": 0.9702, | |
| "pauc10": 0.8351, | |
| "precision": 0.8718, | |
| "f1": 0.8831 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9338, | |
| "recall": 0.8947, | |
| "fpr": 0.051, | |
| "auroc": 0.9702, | |
| "pauc10": 0.8351, | |
| "precision": 0.8718, | |
| "f1": 0.8831 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9046, | |
| "recall": 0.8705, | |
| "fpr": 0.0569, | |
| "auroc": 0.9644, | |
| "pauc10": 0.775, | |
| "precision": 0.9453, | |
| "f1": 0.9064 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9046, | |
| "recall": 0.8705, | |
| "fpr": 0.0569, | |
| "auroc": 0.9644, | |
| "pauc10": 0.775, | |
| "precision": 0.9453, | |
| "f1": 0.9064 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8448, | |
| "recall": 0.7, | |
| "fpr": 0.0, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8929, | |
| "precision": 1.0, | |
| "f1": 0.8235 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8448, | |
| "recall": 0.7, | |
| "fpr": 0.0, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8929, | |
| "precision": 1.0, | |
| "f1": 0.8235 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8448, | |
| "recall": 0.7, | |
| "fpr": 0.0, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8929, | |
| "precision": 1.0, | |
| "f1": 0.8235 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8448, | |
| "recall": 0.7, | |
| "fpr": 0.0, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8929, | |
| "precision": 1.0, | |
| "f1": 0.8235 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7926, | |
| "recall": 0.7926 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7926, | |
| "recall": 0.7926 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.796, | |
| "recall": 0.796 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.796, | |
| "recall": 0.796 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.975, | |
| "recall": 1.0, | |
| "fpr": 0.05, | |
| "auroc": 0.9937, | |
| "pauc10": 0.9368, | |
| "precision": 0.9524, | |
| "f1": 0.9756 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.975, | |
| "recall": 1.0, | |
| "fpr": 0.05, | |
| "auroc": 0.9937, | |
| "pauc10": 0.9368, | |
| "precision": 0.9524, | |
| "f1": 0.9756 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.975, | |
| "recall": 1.0, | |
| "fpr": 0.05, | |
| "auroc": 0.9937, | |
| "pauc10": 0.9368, | |
| "precision": 0.9524, | |
| "f1": 0.9756 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.975, | |
| "recall": 1.0, | |
| "fpr": 0.05, | |
| "auroc": 0.9937, | |
| "pauc10": 0.9368, | |
| "precision": 0.9524, | |
| "f1": 0.9756 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5047, | |
| "recall": 0.4803, | |
| "fpr": 0.0843, | |
| "auroc": 0.7899, | |
| "pauc10": 0.3953, | |
| "precision": 0.9897, | |
| "f1": 0.6468 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5047, | |
| "recall": 0.4803, | |
| "fpr": 0.0843, | |
| "auroc": 0.7899, | |
| "pauc10": 0.3953, | |
| "precision": 0.9897, | |
| "f1": 0.6468 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5047, | |
| "recall": 0.4803, | |
| "fpr": 0.0843, | |
| "auroc": 0.7899, | |
| "pauc10": 0.3953, | |
| "precision": 0.9897, | |
| "f1": 0.6468 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5047, | |
| "recall": 0.4803, | |
| "fpr": 0.0843, | |
| "auroc": 0.7899, | |
| "pauc10": 0.3953, | |
| "precision": 0.9897, | |
| "f1": 0.6468 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "recall": 0.7147, | |
| "fpr": 0.0013, | |
| "auroc": 0.9223, | |
| "pauc10": 0.7861, | |
| "precision": 0.9981, | |
| "f1": 0.8329 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "recall": 0.7147, | |
| "fpr": 0.0013, | |
| "auroc": 0.9223, | |
| "pauc10": 0.7861, | |
| "precision": 0.9981, | |
| "f1": 0.8329 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "recall": 0.7147, | |
| "fpr": 0.0013, | |
| "auroc": 0.9223, | |
| "pauc10": 0.7861, | |
| "precision": 0.9981, | |
| "f1": 0.8329 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "recall": 0.7147, | |
| "fpr": 0.0013, | |
| "auroc": 0.9223, | |
| "pauc10": 0.7861, | |
| "precision": 0.9981, | |
| "f1": 0.8329 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.955, | |
| "recall": 0.955 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.955, | |
| "recall": 0.955 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.955, | |
| "recall": 0.955 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.955, | |
| "recall": 0.955 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9553, | |
| "recall": 0.9455, | |
| "fpr": 0.0308, | |
| "auroc": 0.9914, | |
| "pauc10": 0.9481, | |
| "precision": 0.9774, | |
| "f1": 0.9612 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9553, | |
| "recall": 0.9455, | |
| "fpr": 0.0308, | |
| "auroc": 0.9914, | |
| "pauc10": 0.9481, | |
| "precision": 0.9774, | |
| "f1": 0.9612 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9554, | |
| "recall": 0.9457, | |
| "fpr": 0.0308, | |
| "auroc": 0.9914, | |
| "pauc10": 0.9479, | |
| "precision": 0.9775, | |
| "f1": 0.9613 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9554, | |
| "recall": 0.9457, | |
| "fpr": 0.0308, | |
| "auroc": 0.9914, | |
| "pauc10": 0.9479, | |
| "precision": 0.9775, | |
| "f1": 0.9613 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9481, | |
| "recall": 0.8982, | |
| "fpr": 0.0019, | |
| "auroc": 0.9976, | |
| "pauc10": 0.9785, | |
| "precision": 0.9979, | |
| "f1": 0.9454 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9481, | |
| "recall": 0.8982, | |
| "fpr": 0.0019, | |
| "auroc": 0.9976, | |
| "pauc10": 0.9785, | |
| "precision": 0.9979, | |
| "f1": 0.9454 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9481, | |
| "recall": 0.8982, | |
| "fpr": 0.0019, | |
| "auroc": 0.9976, | |
| "pauc10": 0.9785, | |
| "precision": 0.9979, | |
| "f1": 0.9454 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9481, | |
| "recall": 0.8982, | |
| "fpr": 0.0019, | |
| "auroc": 0.9976, | |
| "pauc10": 0.9785, | |
| "precision": 0.9979, | |
| "f1": 0.9454 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8183, | |
| "recall": 1.0, | |
| "fpr": 0.3736, | |
| "auroc": 0.9919, | |
| "pauc10": 0.9639, | |
| "precision": 0.7388, | |
| "f1": 0.8498 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8183, | |
| "recall": 1.0, | |
| "fpr": 0.3736, | |
| "auroc": 0.9919, | |
| "pauc10": 0.9639, | |
| "precision": 0.7388, | |
| "f1": 0.8498 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8183, | |
| "recall": 1.0, | |
| "fpr": 0.3736, | |
| "auroc": 0.9919, | |
| "pauc10": 0.9639, | |
| "precision": 0.7388, | |
| "f1": 0.8498 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8183, | |
| "recall": 1.0, | |
| "fpr": 0.3736, | |
| "auroc": 0.9919, | |
| "pauc10": 0.9639, | |
| "precision": 0.7388, | |
| "f1": 0.8498 | |
| } | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9617, | |
| "fpr": 0.0383 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9617, | |
| "fpr": 0.0383 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9617, | |
| "fpr": 0.0383 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9617, | |
| "fpr": 0.0383 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9947, | |
| "fpr": 0.0053 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9947, | |
| "fpr": 0.0053 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9947, | |
| "fpr": 0.0053 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9947, | |
| "fpr": 0.0053 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9485, | |
| "recall": 0.8947, | |
| "fpr": 0.0306, | |
| "auroc": 0.978, | |
| "pauc10": 0.8448, | |
| "precision": 0.9189, | |
| "f1": 0.9067 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9485, | |
| "recall": 0.8947, | |
| "fpr": 0.0306, | |
| "auroc": 0.978, | |
| "pauc10": 0.8448, | |
| "precision": 0.9189, | |
| "f1": 0.9067 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9084, | |
| "recall": 0.8633, | |
| "fpr": 0.0407, | |
| "auroc": 0.9696, | |
| "pauc10": 0.7817, | |
| "precision": 0.96, | |
| "f1": 0.9091 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9084, | |
| "recall": 0.8633, | |
| "fpr": 0.0407, | |
| "auroc": 0.9696, | |
| "pauc10": 0.7817, | |
| "precision": 0.96, | |
| "f1": 0.9091 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.9762, | |
| "pauc10": 0.8899, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.9762, | |
| "pauc10": 0.8899, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.9762, | |
| "pauc10": 0.8899, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.9762, | |
| "pauc10": 0.8899, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7582, | |
| "recall": 0.7582 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7582, | |
| "recall": 0.7582 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7626, | |
| "recall": 0.7626 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7626, | |
| "recall": 0.7626 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.975, | |
| "recall": 1.0, | |
| "fpr": 0.05, | |
| "auroc": 0.9937, | |
| "pauc10": 0.9368, | |
| "precision": 0.9524, | |
| "f1": 0.9756 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.975, | |
| "recall": 1.0, | |
| "fpr": 0.05, | |
| "auroc": 0.9937, | |
| "pauc10": 0.9368, | |
| "precision": 0.9524, | |
| "f1": 0.9756 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.975, | |
| "recall": 1.0, | |
| "fpr": 0.05, | |
| "auroc": 0.9937, | |
| "pauc10": 0.9368, | |
| "precision": 0.9524, | |
| "f1": 0.9756 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.975, | |
| "recall": 1.0, | |
| "fpr": 0.05, | |
| "auroc": 0.9937, | |
| "pauc10": 0.9368, | |
| "precision": 0.9524, | |
| "f1": 0.9756 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.393, | |
| "recall": 0.3593, | |
| "fpr": 0.0393, | |
| "auroc": 0.802, | |
| "pauc10": 0.3529, | |
| "precision": 0.9935, | |
| "f1": 0.5278 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.393, | |
| "recall": 0.3593, | |
| "fpr": 0.0393, | |
| "auroc": 0.802, | |
| "pauc10": 0.3529, | |
| "precision": 0.9935, | |
| "f1": 0.5278 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.393, | |
| "recall": 0.3593, | |
| "fpr": 0.0393, | |
| "auroc": 0.802, | |
| "pauc10": 0.3529, | |
| "precision": 0.9935, | |
| "f1": 0.5278 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.393, | |
| "recall": 0.3593, | |
| "fpr": 0.0393, | |
| "auroc": 0.802, | |
| "pauc10": 0.3529, | |
| "precision": 0.9935, | |
| "f1": 0.5278 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9407, | |
| "recall": 0.8813, | |
| "fpr": 0.0, | |
| "auroc": 0.9932, | |
| "pauc10": 0.9731, | |
| "precision": 1.0, | |
| "f1": 0.9369 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9407, | |
| "recall": 0.8813, | |
| "fpr": 0.0, | |
| "auroc": 0.9932, | |
| "pauc10": 0.9731, | |
| "precision": 1.0, | |
| "f1": 0.9369 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9407, | |
| "recall": 0.8813, | |
| "fpr": 0.0, | |
| "auroc": 0.9932, | |
| "pauc10": 0.9731, | |
| "precision": 1.0, | |
| "f1": 0.9369 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9407, | |
| "recall": 0.8813, | |
| "fpr": 0.0, | |
| "auroc": 0.9932, | |
| "pauc10": 0.9731, | |
| "precision": 1.0, | |
| "f1": 0.9369 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.991, | |
| "recall": 0.991 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.991, | |
| "recall": 0.991 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.991, | |
| "recall": 0.991 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.991, | |
| "recall": 0.991 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9543, | |
| "recall": 0.9436, | |
| "fpr": 0.0308, | |
| "auroc": 0.9912, | |
| "pauc10": 0.9462, | |
| "precision": 0.9774, | |
| "f1": 0.9602 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9543, | |
| "recall": 0.9436, | |
| "fpr": 0.0308, | |
| "auroc": 0.9912, | |
| "pauc10": 0.9462, | |
| "precision": 0.9774, | |
| "f1": 0.9602 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9544, | |
| "recall": 0.9438, | |
| "fpr": 0.0308, | |
| "auroc": 0.9912, | |
| "pauc10": 0.9461, | |
| "precision": 0.9775, | |
| "f1": 0.9604 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9544, | |
| "recall": 0.9438, | |
| "fpr": 0.0308, | |
| "auroc": 0.9912, | |
| "pauc10": 0.9461, | |
| "precision": 0.9775, | |
| "f1": 0.9604 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9481, | |
| "recall": 0.8982, | |
| "fpr": 0.0019, | |
| "auroc": 0.9976, | |
| "pauc10": 0.9785, | |
| "precision": 0.9979, | |
| "f1": 0.9454 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9481, | |
| "recall": 0.8982, | |
| "fpr": 0.0019, | |
| "auroc": 0.9976, | |
| "pauc10": 0.9785, | |
| "precision": 0.9979, | |
| "f1": 0.9454 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9481, | |
| "recall": 0.8982, | |
| "fpr": 0.0019, | |
| "auroc": 0.9976, | |
| "pauc10": 0.9785, | |
| "precision": 0.9979, | |
| "f1": 0.9454 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9481, | |
| "recall": 0.8982, | |
| "fpr": 0.0019, | |
| "auroc": 0.9976, | |
| "pauc10": 0.9785, | |
| "precision": 0.9979, | |
| "f1": 0.9454 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8183, | |
| "recall": 1.0, | |
| "fpr": 0.3736, | |
| "auroc": 0.9919, | |
| "pauc10": 0.9639, | |
| "precision": 0.7388, | |
| "f1": 0.8498 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8183, | |
| "recall": 1.0, | |
| "fpr": 0.3736, | |
| "auroc": 0.9919, | |
| "pauc10": 0.9639, | |
| "precision": 0.7388, | |
| "f1": 0.8498 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8183, | |
| "recall": 1.0, | |
| "fpr": 0.3736, | |
| "auroc": 0.9919, | |
| "pauc10": 0.9639, | |
| "precision": 0.7388, | |
| "f1": 0.8498 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8183, | |
| "recall": 1.0, | |
| "fpr": 0.3736, | |
| "auroc": 0.9919, | |
| "pauc10": 0.9639, | |
| "precision": 0.7388, | |
| "f1": 0.8498 | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "horizon-base": { | |
| "serve": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9977, | |
| "fpr": 0.0023 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9977, | |
| "fpr": 0.0023 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9978, | |
| "fpr": 0.0022 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9978, | |
| "fpr": 0.0022 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9931, | |
| "fpr": 0.0069 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9931, | |
| "fpr": 0.0069 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9932, | |
| "fpr": 0.0068 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9932, | |
| "fpr": 0.0068 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9485, | |
| "recall": 0.8947, | |
| "fpr": 0.0306, | |
| "auroc": 0.9737, | |
| "pauc10": 0.8786, | |
| "precision": 0.9189, | |
| "f1": 0.9067 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9485, | |
| "recall": 0.8947, | |
| "fpr": 0.0306, | |
| "auroc": 0.9737, | |
| "pauc10": 0.8786, | |
| "precision": 0.9189, | |
| "f1": 0.9067 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9237, | |
| "recall": 0.8921, | |
| "fpr": 0.0407, | |
| "auroc": 0.9716, | |
| "pauc10": 0.8403, | |
| "precision": 0.9612, | |
| "f1": 0.9254 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9237, | |
| "recall": 0.8921, | |
| "fpr": 0.0407, | |
| "auroc": 0.9716, | |
| "pauc10": 0.8403, | |
| "precision": 0.9612, | |
| "f1": 0.9254 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.819, | |
| "recall": 0.65, | |
| "fpr": 0.0, | |
| "auroc": 0.9554, | |
| "pauc10": 0.8476, | |
| "precision": 1.0, | |
| "f1": 0.7879 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.819, | |
| "recall": 0.65, | |
| "fpr": 0.0, | |
| "auroc": 0.9554, | |
| "pauc10": 0.8476, | |
| "precision": 1.0, | |
| "f1": 0.7879 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.819, | |
| "recall": 0.65, | |
| "fpr": 0.0, | |
| "auroc": 0.9554, | |
| "pauc10": 0.8476, | |
| "precision": 1.0, | |
| "f1": 0.7879 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.819, | |
| "recall": 0.65, | |
| "fpr": 0.0, | |
| "auroc": 0.9554, | |
| "pauc10": 0.8476, | |
| "precision": 1.0, | |
| "f1": 0.7879 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.8292, | |
| "recall": 0.8292 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.8292, | |
| "recall": 0.8292 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8326, | |
| "recall": 0.8326 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8326, | |
| "recall": 0.8326 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9708, | |
| "recall": 0.9917, | |
| "fpr": 0.05, | |
| "auroc": 0.9917, | |
| "pauc10": 0.9215, | |
| "precision": 0.952, | |
| "f1": 0.9714 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9708, | |
| "recall": 0.9917, | |
| "fpr": 0.05, | |
| "auroc": 0.9917, | |
| "pauc10": 0.9215, | |
| "precision": 0.952, | |
| "f1": 0.9714 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9708, | |
| "recall": 0.9917, | |
| "fpr": 0.05, | |
| "auroc": 0.9917, | |
| "pauc10": 0.9215, | |
| "precision": 0.952, | |
| "f1": 0.9714 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9708, | |
| "recall": 0.9917, | |
| "fpr": 0.05, | |
| "auroc": 0.9917, | |
| "pauc10": 0.9215, | |
| "precision": 0.952, | |
| "f1": 0.9714 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.6045, | |
| "recall": 0.583, | |
| "fpr": 0.0337, | |
| "auroc": 0.864, | |
| "pauc10": 0.6243, | |
| "precision": 0.9966, | |
| "f1": 0.7356 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.6045, | |
| "recall": 0.583, | |
| "fpr": 0.0337, | |
| "auroc": 0.864, | |
| "pauc10": 0.6243, | |
| "precision": 0.9966, | |
| "f1": 0.7356 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.6045, | |
| "recall": 0.583, | |
| "fpr": 0.0337, | |
| "auroc": 0.864, | |
| "pauc10": 0.6243, | |
| "precision": 0.9966, | |
| "f1": 0.7356 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.6045, | |
| "recall": 0.583, | |
| "fpr": 0.0337, | |
| "auroc": 0.864, | |
| "pauc10": 0.6243, | |
| "precision": 0.9966, | |
| "f1": 0.7356 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8807, | |
| "recall": 0.7627, | |
| "fpr": 0.0013, | |
| "auroc": 0.9357, | |
| "pauc10": 0.8044, | |
| "precision": 0.9983, | |
| "f1": 0.8647 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8807, | |
| "recall": 0.7627, | |
| "fpr": 0.0013, | |
| "auroc": 0.9357, | |
| "pauc10": 0.8044, | |
| "precision": 0.9983, | |
| "f1": 0.8647 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8807, | |
| "recall": 0.7627, | |
| "fpr": 0.0013, | |
| "auroc": 0.9357, | |
| "pauc10": 0.8044, | |
| "precision": 0.9983, | |
| "f1": 0.8647 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8807, | |
| "recall": 0.7627, | |
| "fpr": 0.0013, | |
| "auroc": 0.9357, | |
| "pauc10": 0.8044, | |
| "precision": 0.9983, | |
| "f1": 0.8647 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9865, | |
| "recall": 0.9865 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9865, | |
| "recall": 0.9865 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9865, | |
| "recall": 0.9865 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9865, | |
| "recall": 0.9865 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9543, | |
| "recall": 0.9418, | |
| "fpr": 0.0282, | |
| "auroc": 0.9914, | |
| "pauc10": 0.9469, | |
| "precision": 0.9792, | |
| "f1": 0.9601 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9543, | |
| "recall": 0.9418, | |
| "fpr": 0.0282, | |
| "auroc": 0.9914, | |
| "pauc10": 0.9469, | |
| "precision": 0.9792, | |
| "f1": 0.9601 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9544, | |
| "recall": 0.942, | |
| "fpr": 0.0282, | |
| "auroc": 0.9914, | |
| "pauc10": 0.9468, | |
| "precision": 0.9793, | |
| "f1": 0.9603 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9544, | |
| "recall": 0.942, | |
| "fpr": 0.0282, | |
| "auroc": 0.9914, | |
| "pauc10": 0.9468, | |
| "precision": 0.9793, | |
| "f1": 0.9603 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.951, | |
| "recall": 0.9049, | |
| "fpr": 0.0029, | |
| "auroc": 0.995, | |
| "pauc10": 0.964, | |
| "precision": 0.9969, | |
| "f1": 0.9486 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.951, | |
| "recall": 0.9049, | |
| "fpr": 0.0029, | |
| "auroc": 0.995, | |
| "pauc10": 0.964, | |
| "precision": 0.9969, | |
| "f1": 0.9486 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.951, | |
| "recall": 0.9049, | |
| "fpr": 0.0029, | |
| "auroc": 0.995, | |
| "pauc10": 0.964, | |
| "precision": 0.9969, | |
| "f1": 0.9486 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.951, | |
| "recall": 0.9049, | |
| "fpr": 0.0029, | |
| "auroc": 0.995, | |
| "pauc10": 0.964, | |
| "precision": 0.9969, | |
| "f1": 0.9486 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.844, | |
| "recall": 1.0, | |
| "fpr": 0.3208, | |
| "auroc": 0.992, | |
| "pauc10": 0.946, | |
| "precision": 0.7671, | |
| "f1": 0.8682 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.844, | |
| "recall": 1.0, | |
| "fpr": 0.3208, | |
| "auroc": 0.992, | |
| "pauc10": 0.946, | |
| "precision": 0.7671, | |
| "f1": 0.8682 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.844, | |
| "recall": 1.0, | |
| "fpr": 0.3208, | |
| "auroc": 0.992, | |
| "pauc10": 0.946, | |
| "precision": 0.7671, | |
| "f1": 0.8682 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.844, | |
| "recall": 1.0, | |
| "fpr": 0.3208, | |
| "auroc": 0.992, | |
| "pauc10": 0.946, | |
| "precision": 0.7671, | |
| "f1": 0.8682 | |
| } | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9977, | |
| "fpr": 0.0023 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9977, | |
| "fpr": 0.0023 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9978, | |
| "fpr": 0.0022 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9978, | |
| "fpr": 0.0022 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9931, | |
| "fpr": 0.0069 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9931, | |
| "fpr": 0.0069 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9932, | |
| "fpr": 0.0068 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9932, | |
| "fpr": 0.0068 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9632, | |
| "recall": 0.8947, | |
| "fpr": 0.0102, | |
| "auroc": 0.9823, | |
| "pauc10": 0.9044, | |
| "precision": 0.9714, | |
| "f1": 0.9315 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9632, | |
| "recall": 0.8947, | |
| "fpr": 0.0102, | |
| "auroc": 0.9823, | |
| "pauc10": 0.9044, | |
| "precision": 0.9714, | |
| "f1": 0.9315 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9275, | |
| "recall": 0.8849, | |
| "fpr": 0.0244, | |
| "auroc": 0.9798, | |
| "pauc10": 0.879, | |
| "precision": 0.9762, | |
| "f1": 0.9283 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9275, | |
| "recall": 0.8849, | |
| "fpr": 0.0244, | |
| "auroc": 0.9798, | |
| "pauc10": 0.879, | |
| "precision": 0.9762, | |
| "f1": 0.9283 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8103, | |
| "recall": 0.6333, | |
| "fpr": 0.0, | |
| "auroc": 0.9548, | |
| "pauc10": 0.8417, | |
| "precision": 1.0, | |
| "f1": 0.7755 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8103, | |
| "recall": 0.6333, | |
| "fpr": 0.0, | |
| "auroc": 0.9548, | |
| "pauc10": 0.8417, | |
| "precision": 1.0, | |
| "f1": 0.7755 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8103, | |
| "recall": 0.6333, | |
| "fpr": 0.0, | |
| "auroc": 0.9548, | |
| "pauc10": 0.8417, | |
| "precision": 1.0, | |
| "f1": 0.7755 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8103, | |
| "recall": 0.6333, | |
| "fpr": 0.0, | |
| "auroc": 0.9548, | |
| "pauc10": 0.8417, | |
| "precision": 1.0, | |
| "f1": 0.7755 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.8187, | |
| "recall": 0.8187 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.8187, | |
| "recall": 0.8187 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8228, | |
| "recall": 0.8228 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8228, | |
| "recall": 0.8228 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9708, | |
| "recall": 0.9917, | |
| "fpr": 0.05, | |
| "auroc": 0.9917, | |
| "pauc10": 0.9208, | |
| "precision": 0.952, | |
| "f1": 0.9714 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9708, | |
| "recall": 0.9917, | |
| "fpr": 0.05, | |
| "auroc": 0.9917, | |
| "pauc10": 0.9208, | |
| "precision": 0.952, | |
| "f1": 0.9714 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9708, | |
| "recall": 0.9917, | |
| "fpr": 0.05, | |
| "auroc": 0.9917, | |
| "pauc10": 0.9208, | |
| "precision": 0.952, | |
| "f1": 0.9714 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9708, | |
| "recall": 0.9917, | |
| "fpr": 0.05, | |
| "auroc": 0.9917, | |
| "pauc10": 0.9208, | |
| "precision": 0.952, | |
| "f1": 0.9714 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.4896, | |
| "recall": 0.4607, | |
| "fpr": 0.0225, | |
| "auroc": 0.8363, | |
| "pauc10": 0.5324, | |
| "precision": 0.9971, | |
| "f1": 0.6302 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.4896, | |
| "recall": 0.4607, | |
| "fpr": 0.0225, | |
| "auroc": 0.8363, | |
| "pauc10": 0.5324, | |
| "precision": 0.9971, | |
| "f1": 0.6302 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.4896, | |
| "recall": 0.4607, | |
| "fpr": 0.0225, | |
| "auroc": 0.8363, | |
| "pauc10": 0.5324, | |
| "precision": 0.9971, | |
| "f1": 0.6302 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.4896, | |
| "recall": 0.4607, | |
| "fpr": 0.0225, | |
| "auroc": 0.8363, | |
| "pauc10": 0.5324, | |
| "precision": 0.9971, | |
| "f1": 0.6302 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9707, | |
| "recall": 0.9413, | |
| "fpr": 0.0, | |
| "auroc": 0.9969, | |
| "pauc10": 0.9825, | |
| "precision": 1.0, | |
| "f1": 0.9698 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9707, | |
| "recall": 0.9413, | |
| "fpr": 0.0, | |
| "auroc": 0.9969, | |
| "pauc10": 0.9825, | |
| "precision": 1.0, | |
| "f1": 0.9698 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9707, | |
| "recall": 0.9413, | |
| "fpr": 0.0, | |
| "auroc": 0.9969, | |
| "pauc10": 0.9825, | |
| "precision": 1.0, | |
| "f1": 0.9698 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9707, | |
| "recall": 0.9413, | |
| "fpr": 0.0, | |
| "auroc": 0.9969, | |
| "pauc10": 0.9825, | |
| "precision": 1.0, | |
| "f1": 0.9698 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9975, | |
| "recall": 0.9975 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9975, | |
| "recall": 0.9975 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9975, | |
| "recall": 0.9975 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9975, | |
| "recall": 0.9975 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9553, | |
| "recall": 0.9436, | |
| "fpr": 0.0282, | |
| "auroc": 0.9912, | |
| "pauc10": 0.9452, | |
| "precision": 0.9792, | |
| "f1": 0.9611 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9553, | |
| "recall": 0.9436, | |
| "fpr": 0.0282, | |
| "auroc": 0.9912, | |
| "pauc10": 0.9452, | |
| "precision": 0.9792, | |
| "f1": 0.9611 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9554, | |
| "recall": 0.9438, | |
| "fpr": 0.0282, | |
| "auroc": 0.9912, | |
| "pauc10": 0.9451, | |
| "precision": 0.9793, | |
| "f1": 0.9613 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9554, | |
| "recall": 0.9438, | |
| "fpr": 0.0282, | |
| "auroc": 0.9912, | |
| "pauc10": 0.9451, | |
| "precision": 0.9793, | |
| "f1": 0.9613 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.951, | |
| "recall": 0.9049, | |
| "fpr": 0.0029, | |
| "auroc": 0.995, | |
| "pauc10": 0.964, | |
| "precision": 0.9969, | |
| "f1": 0.9486 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.951, | |
| "recall": 0.9049, | |
| "fpr": 0.0029, | |
| "auroc": 0.995, | |
| "pauc10": 0.964, | |
| "precision": 0.9969, | |
| "f1": 0.9486 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.951, | |
| "recall": 0.9049, | |
| "fpr": 0.0029, | |
| "auroc": 0.995, | |
| "pauc10": 0.964, | |
| "precision": 0.9969, | |
| "f1": 0.9486 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.951, | |
| "recall": 0.9049, | |
| "fpr": 0.0029, | |
| "auroc": 0.995, | |
| "pauc10": 0.964, | |
| "precision": 0.9969, | |
| "f1": 0.9486 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.844, | |
| "recall": 1.0, | |
| "fpr": 0.3208, | |
| "auroc": 0.992, | |
| "pauc10": 0.946, | |
| "precision": 0.7671, | |
| "f1": 0.8682 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.844, | |
| "recall": 1.0, | |
| "fpr": 0.3208, | |
| "auroc": 0.992, | |
| "pauc10": 0.946, | |
| "precision": 0.7671, | |
| "f1": 0.8682 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.844, | |
| "recall": 1.0, | |
| "fpr": 0.3208, | |
| "auroc": 0.992, | |
| "pauc10": 0.946, | |
| "precision": 0.7671, | |
| "f1": 0.8682 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.844, | |
| "recall": 1.0, | |
| "fpr": 0.3208, | |
| "auroc": 0.992, | |
| "pauc10": 0.946, | |
| "precision": 0.7671, | |
| "f1": 0.8682 | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "horizon-large": { | |
| "serve": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9322, | |
| "fpr": 0.0678 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9322, | |
| "fpr": 0.0678 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9322, | |
| "fpr": 0.0678 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9322, | |
| "fpr": 0.0678 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9939, | |
| "fpr": 0.0061 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9939, | |
| "fpr": 0.0061 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9932, | |
| "fpr": 0.0068 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9932, | |
| "fpr": 0.0068 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9338, | |
| "recall": 0.8158, | |
| "fpr": 0.0204, | |
| "auroc": 0.9656, | |
| "pauc10": 0.8432, | |
| "precision": 0.9394, | |
| "f1": 0.8732 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9338, | |
| "recall": 0.8158, | |
| "fpr": 0.0204, | |
| "auroc": 0.9656, | |
| "pauc10": 0.8432, | |
| "precision": 0.9394, | |
| "f1": 0.8732 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.8817, | |
| "recall": 0.8058, | |
| "fpr": 0.0325, | |
| "auroc": 0.9566, | |
| "pauc10": 0.7439, | |
| "precision": 0.9655, | |
| "f1": 0.8784 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.8817, | |
| "recall": 0.8058, | |
| "fpr": 0.0325, | |
| "auroc": 0.9566, | |
| "pauc10": 0.7439, | |
| "precision": 0.9655, | |
| "f1": 0.8784 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.95, | |
| "pauc10": 0.8667, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.95, | |
| "pauc10": 0.8667, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.95, | |
| "pauc10": 0.8667, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.95, | |
| "pauc10": 0.8667, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7638, | |
| "recall": 0.7638 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7638, | |
| "recall": 0.7638 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7671, | |
| "recall": 0.7671 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7671, | |
| "recall": 0.7671 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9917, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 1.0, | |
| "pauc10": 1.0, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9917, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 1.0, | |
| "pauc10": 1.0, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9917, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 1.0, | |
| "pauc10": 1.0, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9917, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 1.0, | |
| "pauc10": 1.0, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5041, | |
| "recall": 0.4783, | |
| "fpr": 0.0618, | |
| "auroc": 0.8249, | |
| "pauc10": 0.4313, | |
| "precision": 0.9924, | |
| "f1": 0.6455 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5041, | |
| "recall": 0.4783, | |
| "fpr": 0.0618, | |
| "auroc": 0.8249, | |
| "pauc10": 0.4313, | |
| "precision": 0.9924, | |
| "f1": 0.6455 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5041, | |
| "recall": 0.4783, | |
| "fpr": 0.0618, | |
| "auroc": 0.8249, | |
| "pauc10": 0.4313, | |
| "precision": 0.9924, | |
| "f1": 0.6455 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5041, | |
| "recall": 0.4783, | |
| "fpr": 0.0618, | |
| "auroc": 0.8249, | |
| "pauc10": 0.4313, | |
| "precision": 0.9924, | |
| "f1": 0.6455 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8573, | |
| "recall": 0.72, | |
| "fpr": 0.0053, | |
| "auroc": 0.9288, | |
| "pauc10": 0.7808, | |
| "precision": 0.9926, | |
| "f1": 0.8346 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8573, | |
| "recall": 0.72, | |
| "fpr": 0.0053, | |
| "auroc": 0.9288, | |
| "pauc10": 0.7808, | |
| "precision": 0.9926, | |
| "f1": 0.8346 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8573, | |
| "recall": 0.72, | |
| "fpr": 0.0053, | |
| "auroc": 0.9288, | |
| "pauc10": 0.7808, | |
| "precision": 0.9926, | |
| "f1": 0.8346 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.8573, | |
| "recall": 0.72, | |
| "fpr": 0.0053, | |
| "auroc": 0.9288, | |
| "pauc10": 0.7808, | |
| "precision": 0.9926, | |
| "f1": 0.8346 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.992, | |
| "recall": 0.992 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.992, | |
| "recall": 0.992 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.992, | |
| "recall": 0.992 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.992, | |
| "recall": 0.992 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.95, | |
| "recall": 0.9364, | |
| "fpr": 0.0308, | |
| "auroc": 0.9888, | |
| "pauc10": 0.9196, | |
| "precision": 0.9772, | |
| "f1": 0.9564 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.95, | |
| "recall": 0.9364, | |
| "fpr": 0.0308, | |
| "auroc": 0.9888, | |
| "pauc10": 0.9196, | |
| "precision": 0.9772, | |
| "f1": 0.9564 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9501, | |
| "recall": 0.9366, | |
| "fpr": 0.0308, | |
| "auroc": 0.9888, | |
| "pauc10": 0.9195, | |
| "precision": 0.9773, | |
| "f1": 0.9565 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9501, | |
| "recall": 0.9366, | |
| "fpr": 0.0308, | |
| "auroc": 0.9888, | |
| "pauc10": 0.9195, | |
| "precision": 0.9773, | |
| "f1": 0.9565 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "recall": 0.8754, | |
| "fpr": 0.0057, | |
| "auroc": 0.994, | |
| "pauc10": 0.9471, | |
| "precision": 0.9935, | |
| "f1": 0.9307 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "recall": 0.8754, | |
| "fpr": 0.0057, | |
| "auroc": 0.994, | |
| "pauc10": 0.9471, | |
| "precision": 0.9935, | |
| "f1": 0.9307 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "recall": 0.8754, | |
| "fpr": 0.0057, | |
| "auroc": 0.994, | |
| "pauc10": 0.9471, | |
| "precision": 0.9935, | |
| "f1": 0.9307 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "recall": 0.8754, | |
| "fpr": 0.0057, | |
| "auroc": 0.994, | |
| "pauc10": 0.9471, | |
| "precision": 0.9935, | |
| "f1": 0.9307 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8826, | |
| "recall": 1.0, | |
| "fpr": 0.2415, | |
| "auroc": 0.9992, | |
| "pauc10": 0.9948, | |
| "precision": 0.814, | |
| "f1": 0.8974 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8826, | |
| "recall": 1.0, | |
| "fpr": 0.2415, | |
| "auroc": 0.9992, | |
| "pauc10": 0.9948, | |
| "precision": 0.814, | |
| "f1": 0.8974 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8826, | |
| "recall": 1.0, | |
| "fpr": 0.2415, | |
| "auroc": 0.9992, | |
| "pauc10": 0.9948, | |
| "precision": 0.814, | |
| "f1": 0.8974 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8826, | |
| "recall": 1.0, | |
| "fpr": 0.2415, | |
| "auroc": 0.9992, | |
| "pauc10": 0.9948, | |
| "precision": 0.814, | |
| "f1": 0.8974 | |
| } | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9322, | |
| "fpr": 0.0678 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9322, | |
| "fpr": 0.0678 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9322, | |
| "fpr": 0.0678 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9322, | |
| "fpr": 0.0678 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9939, | |
| "fpr": 0.0061 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9939, | |
| "fpr": 0.0061 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9932, | |
| "fpr": 0.0068 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9932, | |
| "fpr": 0.0068 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9338, | |
| "recall": 0.8158, | |
| "fpr": 0.0204, | |
| "auroc": 0.9643, | |
| "pauc10": 0.8298, | |
| "precision": 0.9394, | |
| "f1": 0.8732 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9338, | |
| "recall": 0.8158, | |
| "fpr": 0.0204, | |
| "auroc": 0.9643, | |
| "pauc10": 0.8298, | |
| "precision": 0.9394, | |
| "f1": 0.8732 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.874, | |
| "recall": 0.7914, | |
| "fpr": 0.0325, | |
| "auroc": 0.9579, | |
| "pauc10": 0.7462, | |
| "precision": 0.9649, | |
| "f1": 0.8696 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.874, | |
| "recall": 0.7914, | |
| "fpr": 0.0325, | |
| "auroc": 0.9579, | |
| "pauc10": 0.7462, | |
| "precision": 0.9649, | |
| "f1": 0.8696 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.95, | |
| "pauc10": 0.8667, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.95, | |
| "pauc10": 0.8667, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.95, | |
| "pauc10": 0.8667, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8362, | |
| "recall": 0.6833, | |
| "fpr": 0.0, | |
| "auroc": 0.95, | |
| "pauc10": 0.8667, | |
| "precision": 1.0, | |
| "f1": 0.8119 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7638, | |
| "recall": 0.7638 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7638, | |
| "recall": 0.7638 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7655, | |
| "recall": 0.7655 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7655, | |
| "recall": 0.7655 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9917, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 1.0, | |
| "pauc10": 1.0, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9917, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 1.0, | |
| "pauc10": 1.0, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9917, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 1.0, | |
| "pauc10": 1.0, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.9917, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 1.0, | |
| "pauc10": 1.0, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5419, | |
| "recall": 0.5183, | |
| "fpr": 0.0618, | |
| "auroc": 0.8601, | |
| "pauc10": 0.4691, | |
| "precision": 0.993, | |
| "f1": 0.6811 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5419, | |
| "recall": 0.5183, | |
| "fpr": 0.0618, | |
| "auroc": 0.8601, | |
| "pauc10": 0.4691, | |
| "precision": 0.993, | |
| "f1": 0.6811 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5419, | |
| "recall": 0.5183, | |
| "fpr": 0.0618, | |
| "auroc": 0.8601, | |
| "pauc10": 0.4691, | |
| "precision": 0.993, | |
| "f1": 0.6811 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.5419, | |
| "recall": 0.5183, | |
| "fpr": 0.0618, | |
| "auroc": 0.8601, | |
| "pauc10": 0.4691, | |
| "precision": 0.993, | |
| "f1": 0.6811 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9513, | |
| "recall": 0.9067, | |
| "fpr": 0.004, | |
| "auroc": 0.9865, | |
| "pauc10": 0.9539, | |
| "precision": 0.9956, | |
| "f1": 0.9491 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9513, | |
| "recall": 0.9067, | |
| "fpr": 0.004, | |
| "auroc": 0.9865, | |
| "pauc10": 0.9539, | |
| "precision": 0.9956, | |
| "f1": 0.9491 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9513, | |
| "recall": 0.9067, | |
| "fpr": 0.004, | |
| "auroc": 0.9865, | |
| "pauc10": 0.9539, | |
| "precision": 0.9956, | |
| "f1": 0.9491 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.9513, | |
| "recall": 0.9067, | |
| "fpr": 0.004, | |
| "auroc": 0.9865, | |
| "pauc10": 0.9539, | |
| "precision": 0.9956, | |
| "f1": 0.9491 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9955, | |
| "recall": 0.9955 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9955, | |
| "recall": 0.9955 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9955, | |
| "recall": 0.9955 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.9955, | |
| "recall": 0.9955 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9511, | |
| "recall": 0.9382, | |
| "fpr": 0.0308, | |
| "auroc": 0.9888, | |
| "pauc10": 0.9199, | |
| "precision": 0.9773, | |
| "f1": 0.9573 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9511, | |
| "recall": 0.9382, | |
| "fpr": 0.0308, | |
| "auroc": 0.9888, | |
| "pauc10": 0.9199, | |
| "precision": 0.9773, | |
| "f1": 0.9573 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9512, | |
| "recall": 0.9384, | |
| "fpr": 0.0308, | |
| "auroc": 0.9888, | |
| "pauc10": 0.9198, | |
| "precision": 0.9774, | |
| "f1": 0.9575 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9512, | |
| "recall": 0.9384, | |
| "fpr": 0.0308, | |
| "auroc": 0.9888, | |
| "pauc10": 0.9198, | |
| "precision": 0.9774, | |
| "f1": 0.9575 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "recall": 0.8754, | |
| "fpr": 0.0057, | |
| "auroc": 0.994, | |
| "pauc10": 0.9471, | |
| "precision": 0.9935, | |
| "f1": 0.9307 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "recall": 0.8754, | |
| "fpr": 0.0057, | |
| "auroc": 0.994, | |
| "pauc10": 0.9471, | |
| "precision": 0.9935, | |
| "f1": 0.9307 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "recall": 0.8754, | |
| "fpr": 0.0057, | |
| "auroc": 0.994, | |
| "pauc10": 0.9471, | |
| "precision": 0.9935, | |
| "f1": 0.9307 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "recall": 0.8754, | |
| "fpr": 0.0057, | |
| "auroc": 0.994, | |
| "pauc10": 0.9471, | |
| "precision": 0.9935, | |
| "f1": 0.9307 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8826, | |
| "recall": 1.0, | |
| "fpr": 0.2415, | |
| "auroc": 0.9992, | |
| "pauc10": 0.9948, | |
| "precision": 0.814, | |
| "f1": 0.8974 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8826, | |
| "recall": 1.0, | |
| "fpr": 0.2415, | |
| "auroc": 0.9992, | |
| "pauc10": 0.9948, | |
| "precision": 0.814, | |
| "f1": 0.8974 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8826, | |
| "recall": 1.0, | |
| "fpr": 0.2415, | |
| "auroc": 0.9992, | |
| "pauc10": 0.9948, | |
| "precision": 0.814, | |
| "f1": 0.8974 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.8826, | |
| "recall": 1.0, | |
| "fpr": 0.2415, | |
| "auroc": 0.9992, | |
| "pauc10": 0.9948, | |
| "precision": 0.814, | |
| "f1": 0.8974 | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "galvanize-60m": { | |
| "serve": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.5369, | |
| "fpr": 0.4631 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.4543, | |
| "fpr": 0.5457 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.5369, | |
| "fpr": 0.4631 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.4543, | |
| "fpr": 0.5457 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.2698, | |
| "fpr": 0.7302 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2086, | |
| "fpr": 0.7914 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.2756, | |
| "fpr": 0.7244 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2111, | |
| "fpr": 0.7889 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.2027, | |
| "fpr": 0.7973 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.1418, | |
| "fpr": 0.8582 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.2024, | |
| "fpr": 0.7976 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.1418, | |
| "fpr": 0.8582 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.95, | |
| "acc": 0.9191, | |
| "recall": 0.9474, | |
| "fpr": 0.0918, | |
| "auroc": 0.9643, | |
| "pauc10": 0.9017, | |
| "precision": 0.8, | |
| "f1": 0.8675 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.875, | |
| "recall": 0.9474, | |
| "fpr": 0.1531, | |
| "auroc": 0.9643, | |
| "pauc10": 0.9017, | |
| "precision": 0.7059, | |
| "f1": 0.809 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.95, | |
| "acc": 0.9427, | |
| "recall": 0.964, | |
| "fpr": 0.0813, | |
| "auroc": 0.9747, | |
| "pauc10": 0.9471, | |
| "precision": 0.9306, | |
| "f1": 0.947 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9084, | |
| "recall": 0.964, | |
| "fpr": 0.1545, | |
| "auroc": 0.9747, | |
| "pauc10": 0.9471, | |
| "precision": 0.8758, | |
| "f1": 0.9178 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.95, | |
| "acc": 0.9138, | |
| "recall": 0.9667, | |
| "fpr": 0.1429, | |
| "auroc": 0.9667, | |
| "pauc10": 0.7363, | |
| "precision": 0.8788, | |
| "f1": 0.9206 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8707, | |
| "recall": 0.9833, | |
| "fpr": 0.25, | |
| "auroc": 0.9667, | |
| "pauc10": 0.7363, | |
| "precision": 0.8082, | |
| "f1": 0.8872 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.95, | |
| "acc": 0.9138, | |
| "recall": 0.9667, | |
| "fpr": 0.1429, | |
| "auroc": 0.9667, | |
| "pauc10": 0.7363, | |
| "precision": 0.8788, | |
| "f1": 0.9206 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8707, | |
| "recall": 0.9833, | |
| "fpr": 0.25, | |
| "auroc": 0.9667, | |
| "pauc10": 0.7363, | |
| "precision": 0.8082, | |
| "f1": 0.8872 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.95, | |
| "acc": 0.8048, | |
| "recall": 0.8048 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.8684, | |
| "recall": 0.8684 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.95, | |
| "acc": 0.8133, | |
| "recall": 0.8133 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8739, | |
| "recall": 0.8739 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.95, | |
| "acc": 0.5583, | |
| "recall": 0.875, | |
| "fpr": 0.7583, | |
| "auroc": 0.7026, | |
| "pauc10": 0.2042, | |
| "precision": 0.5357, | |
| "f1": 0.6646 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5417, | |
| "recall": 0.925, | |
| "fpr": 0.8417, | |
| "auroc": 0.7026, | |
| "pauc10": 0.2042, | |
| "precision": 0.5236, | |
| "f1": 0.6687 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.95, | |
| "acc": 0.5583, | |
| "recall": 0.875, | |
| "fpr": 0.7583, | |
| "auroc": 0.7026, | |
| "pauc10": 0.2042, | |
| "precision": 0.5357, | |
| "f1": 0.6646 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5417, | |
| "recall": 0.925, | |
| "fpr": 0.8417, | |
| "auroc": 0.7026, | |
| "pauc10": 0.2042, | |
| "precision": 0.5236, | |
| "f1": 0.6687 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.95, | |
| "acc": 0.2524, | |
| "recall": 0.2157, | |
| "fpr": 0.1292, | |
| "auroc": 0.5769, | |
| "pauc10": 0.0731, | |
| "precision": 0.9657, | |
| "f1": 0.3526 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.3524, | |
| "recall": 0.326, | |
| "fpr": 0.2022, | |
| "auroc": 0.5769, | |
| "pauc10": 0.0731, | |
| "precision": 0.9645, | |
| "f1": 0.4873 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.95, | |
| "acc": 0.2524, | |
| "recall": 0.2157, | |
| "fpr": 0.1292, | |
| "auroc": 0.5769, | |
| "pauc10": 0.0731, | |
| "precision": 0.9657, | |
| "f1": 0.3526 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.3524, | |
| "recall": 0.326, | |
| "fpr": 0.2022, | |
| "auroc": 0.5769, | |
| "pauc10": 0.0731, | |
| "precision": 0.9645, | |
| "f1": 0.4873 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.95, | |
| "acc": 0.5187, | |
| "recall": 0.0507, | |
| "fpr": 0.0133, | |
| "auroc": 0.595, | |
| "pauc10": 0.1151, | |
| "precision": 0.7917, | |
| "f1": 0.0952 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5333, | |
| "recall": 0.104, | |
| "fpr": 0.0373, | |
| "auroc": 0.595, | |
| "pauc10": 0.1151, | |
| "precision": 0.7358, | |
| "f1": 0.1822 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.95, | |
| "acc": 0.5187, | |
| "recall": 0.0507, | |
| "fpr": 0.0133, | |
| "auroc": 0.595, | |
| "pauc10": 0.1151, | |
| "precision": 0.7917, | |
| "f1": 0.0952 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5333, | |
| "recall": 0.104, | |
| "fpr": 0.0373, | |
| "auroc": 0.595, | |
| "pauc10": 0.1151, | |
| "precision": 0.7358, | |
| "f1": 0.1822 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.95, | |
| "acc": 0.4985, | |
| "recall": 0.4985 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.5965, | |
| "recall": 0.5965 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.95, | |
| "acc": 0.4985, | |
| "recall": 0.4985 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.5965, | |
| "recall": 0.5965 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.95, | |
| "acc": 0.9479, | |
| "recall": 0.9855, | |
| "fpr": 0.1051, | |
| "auroc": 0.9799, | |
| "pauc10": 0.8212, | |
| "precision": 0.9297, | |
| "f1": 0.9568 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9372, | |
| "recall": 0.9909, | |
| "fpr": 0.1385, | |
| "auroc": 0.9799, | |
| "pauc10": 0.8212, | |
| "precision": 0.9098, | |
| "f1": 0.9487 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.95, | |
| "acc": 0.948, | |
| "recall": 0.9855, | |
| "fpr": 0.1051, | |
| "auroc": 0.9799, | |
| "pauc10": 0.8218, | |
| "precision": 0.9299, | |
| "f1": 0.9569 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9374, | |
| "recall": 0.9909, | |
| "fpr": 0.1385, | |
| "auroc": 0.9799, | |
| "pauc10": 0.8218, | |
| "precision": 0.9101, | |
| "f1": 0.9488 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.95, | |
| "acc": 0.7887, | |
| "recall": 0.8154, | |
| "fpr": 0.2381, | |
| "auroc": 0.8761, | |
| "pauc10": 0.5132, | |
| "precision": 0.7742, | |
| "f1": 0.7943 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7711, | |
| "recall": 0.8877, | |
| "fpr": 0.3457, | |
| "auroc": 0.8761, | |
| "pauc10": 0.5132, | |
| "precision": 0.7199, | |
| "f1": 0.7951 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.95, | |
| "acc": 0.7887, | |
| "recall": 0.8154, | |
| "fpr": 0.2381, | |
| "auroc": 0.8761, | |
| "pauc10": 0.5132, | |
| "precision": 0.7742, | |
| "f1": 0.7943 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7711, | |
| "recall": 0.8877, | |
| "fpr": 0.3457, | |
| "auroc": 0.8761, | |
| "pauc10": 0.5132, | |
| "precision": 0.7199, | |
| "f1": 0.7951 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.95, | |
| "acc": 0.6422, | |
| "recall": 0.9286, | |
| "fpr": 0.6604, | |
| "auroc": 0.708, | |
| "pauc10": 0.0673, | |
| "precision": 0.5977, | |
| "f1": 0.7273 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.6183, | |
| "recall": 0.975, | |
| "fpr": 0.7585, | |
| "auroc": 0.708, | |
| "pauc10": 0.0673, | |
| "precision": 0.5759, | |
| "f1": 0.7241 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.95, | |
| "acc": 0.6422, | |
| "recall": 0.9286, | |
| "fpr": 0.6604, | |
| "auroc": 0.708, | |
| "pauc10": 0.0673, | |
| "precision": 0.5977, | |
| "f1": 0.7273 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.6183, | |
| "recall": 0.975, | |
| "fpr": 0.7585, | |
| "auroc": 0.708, | |
| "pauc10": 0.0673, | |
| "precision": 0.5759, | |
| "f1": 0.7241 | |
| } | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.5369, | |
| "fpr": 0.4631 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.4543, | |
| "fpr": 0.5457 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.5369, | |
| "fpr": 0.4631 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.4543, | |
| "fpr": 0.5457 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.2698, | |
| "fpr": 0.7302 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2086, | |
| "fpr": 0.7914 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.2756, | |
| "fpr": 0.7244 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2111, | |
| "fpr": 0.7889 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.2027, | |
| "fpr": 0.7973 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.1418, | |
| "fpr": 0.8582 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.95, | |
| "acc": 0.2024, | |
| "fpr": 0.7976 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.1418, | |
| "fpr": 0.8582 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.95, | |
| "acc": 0.9118, | |
| "recall": 0.9474, | |
| "fpr": 0.102, | |
| "auroc": 0.9654, | |
| "pauc10": 0.9017, | |
| "precision": 0.7826, | |
| "f1": 0.8571 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.8824, | |
| "recall": 0.9474, | |
| "fpr": 0.1429, | |
| "auroc": 0.9654, | |
| "pauc10": 0.9017, | |
| "precision": 0.72, | |
| "f1": 0.8182 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.95, | |
| "acc": 0.9427, | |
| "recall": 0.9712, | |
| "fpr": 0.0894, | |
| "auroc": 0.9868, | |
| "pauc10": 0.9537, | |
| "precision": 0.9247, | |
| "f1": 0.9474 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9198, | |
| "recall": 0.9784, | |
| "fpr": 0.1463, | |
| "auroc": 0.9868, | |
| "pauc10": 0.9537, | |
| "precision": 0.8831, | |
| "f1": 0.9283 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.95, | |
| "acc": 0.9224, | |
| "recall": 0.9833, | |
| "fpr": 0.1429, | |
| "auroc": 0.9711, | |
| "pauc10": 0.7399, | |
| "precision": 0.8806, | |
| "f1": 0.9291 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8793, | |
| "recall": 1.0, | |
| "fpr": 0.25, | |
| "auroc": 0.9711, | |
| "pauc10": 0.7399, | |
| "precision": 0.8108, | |
| "f1": 0.8955 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.95, | |
| "acc": 0.9224, | |
| "recall": 0.9833, | |
| "fpr": 0.1429, | |
| "auroc": 0.9711, | |
| "pauc10": 0.7399, | |
| "precision": 0.8806, | |
| "f1": 0.9291 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.8793, | |
| "recall": 1.0, | |
| "fpr": 0.25, | |
| "auroc": 0.9711, | |
| "pauc10": 0.7399, | |
| "precision": 0.8108, | |
| "f1": 0.8955 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.95, | |
| "acc": 0.8501, | |
| "recall": 0.8501 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.8941, | |
| "recall": 0.8941 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.95, | |
| "acc": 0.8574, | |
| "recall": 0.8574 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.899, | |
| "recall": 0.899 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.95, | |
| "acc": 0.5583, | |
| "recall": 0.875, | |
| "fpr": 0.7583, | |
| "auroc": 0.6997, | |
| "pauc10": 0.2028, | |
| "precision": 0.5357, | |
| "f1": 0.6646 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5417, | |
| "recall": 0.925, | |
| "fpr": 0.8417, | |
| "auroc": 0.6997, | |
| "pauc10": 0.2028, | |
| "precision": 0.5236, | |
| "f1": 0.6687 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.95, | |
| "acc": 0.5583, | |
| "recall": 0.875, | |
| "fpr": 0.7583, | |
| "auroc": 0.6997, | |
| "pauc10": 0.2028, | |
| "precision": 0.5357, | |
| "f1": 0.6646 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5417, | |
| "recall": 0.925, | |
| "fpr": 0.8417, | |
| "auroc": 0.6997, | |
| "pauc10": 0.2028, | |
| "precision": 0.5236, | |
| "f1": 0.6687 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.95, | |
| "acc": 0.5063, | |
| "recall": 0.4983, | |
| "fpr": 0.3596, | |
| "auroc": 0.5876, | |
| "pauc10": 0.0588, | |
| "precision": 0.9589, | |
| "f1": 0.6558 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.6152, | |
| "recall": 0.6207, | |
| "fpr": 0.4775, | |
| "auroc": 0.5876, | |
| "pauc10": 0.0588, | |
| "precision": 0.9563, | |
| "f1": 0.7528 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.95, | |
| "acc": 0.5063, | |
| "recall": 0.4983, | |
| "fpr": 0.3596, | |
| "auroc": 0.5876, | |
| "pauc10": 0.0588, | |
| "precision": 0.9589, | |
| "f1": 0.6558 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.6152, | |
| "recall": 0.6207, | |
| "fpr": 0.4775, | |
| "auroc": 0.5876, | |
| "pauc10": 0.0588, | |
| "precision": 0.9563, | |
| "f1": 0.7528 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.95, | |
| "acc": 0.5607, | |
| "recall": 0.1667, | |
| "fpr": 0.0453, | |
| "auroc": 0.6859, | |
| "pauc10": 0.1668, | |
| "precision": 0.7862, | |
| "f1": 0.275 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.592, | |
| "recall": 0.28, | |
| "fpr": 0.096, | |
| "auroc": 0.6859, | |
| "pauc10": 0.1668, | |
| "precision": 0.7447, | |
| "f1": 0.407 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.95, | |
| "acc": 0.5607, | |
| "recall": 0.1667, | |
| "fpr": 0.0453, | |
| "auroc": 0.6859, | |
| "pauc10": 0.1668, | |
| "precision": 0.7862, | |
| "f1": 0.275 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.592, | |
| "recall": 0.28, | |
| "fpr": 0.096, | |
| "auroc": 0.6859, | |
| "pauc10": 0.1668, | |
| "precision": 0.7447, | |
| "f1": 0.407 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.95, | |
| "acc": 0.7125, | |
| "recall": 0.7125 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.7775, | |
| "recall": 0.7775 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.95, | |
| "acc": 0.7125, | |
| "recall": 0.7125 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.7775, | |
| "recall": 0.7775 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.95, | |
| "acc": 0.9479, | |
| "recall": 0.9855, | |
| "fpr": 0.1051, | |
| "auroc": 0.9813, | |
| "pauc10": 0.836, | |
| "precision": 0.9297, | |
| "f1": 0.9568 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9372, | |
| "recall": 0.9909, | |
| "fpr": 0.1385, | |
| "auroc": 0.9813, | |
| "pauc10": 0.836, | |
| "precision": 0.9098, | |
| "f1": 0.9487 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.95, | |
| "acc": 0.948, | |
| "recall": 0.9855, | |
| "fpr": 0.1051, | |
| "auroc": 0.9814, | |
| "pauc10": 0.8366, | |
| "precision": 0.9299, | |
| "f1": 0.9569 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9374, | |
| "recall": 0.9909, | |
| "fpr": 0.1385, | |
| "auroc": 0.9814, | |
| "pauc10": 0.8366, | |
| "precision": 0.9101, | |
| "f1": 0.9488 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.95, | |
| "acc": 0.7887, | |
| "recall": 0.8154, | |
| "fpr": 0.2381, | |
| "auroc": 0.8761, | |
| "pauc10": 0.5132, | |
| "precision": 0.7742, | |
| "f1": 0.7943 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7711, | |
| "recall": 0.8877, | |
| "fpr": 0.3457, | |
| "auroc": 0.8761, | |
| "pauc10": 0.5132, | |
| "precision": 0.7199, | |
| "f1": 0.7951 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.95, | |
| "acc": 0.7887, | |
| "recall": 0.8154, | |
| "fpr": 0.2381, | |
| "auroc": 0.8761, | |
| "pauc10": 0.5132, | |
| "precision": 0.7742, | |
| "f1": 0.7943 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7711, | |
| "recall": 0.8877, | |
| "fpr": 0.3457, | |
| "auroc": 0.8761, | |
| "pauc10": 0.5132, | |
| "precision": 0.7199, | |
| "f1": 0.7951 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.95, | |
| "acc": 0.6422, | |
| "recall": 0.9286, | |
| "fpr": 0.6604, | |
| "auroc": 0.708, | |
| "pauc10": 0.0673, | |
| "precision": 0.5977, | |
| "f1": 0.7273 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.6183, | |
| "recall": 0.975, | |
| "fpr": 0.7585, | |
| "auroc": 0.708, | |
| "pauc10": 0.0673, | |
| "precision": 0.5759, | |
| "f1": 0.7241 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.95, | |
| "acc": 0.6422, | |
| "recall": 0.9286, | |
| "fpr": 0.6604, | |
| "auroc": 0.708, | |
| "pauc10": 0.0673, | |
| "precision": 0.5977, | |
| "f1": 0.7273 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.6183, | |
| "recall": 0.975, | |
| "fpr": 0.7585, | |
| "auroc": 0.708, | |
| "pauc10": 0.0673, | |
| "precision": 0.5759, | |
| "f1": 0.7241 | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "protectai-v2": { | |
| "serve": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.5664, | |
| "fpr": 0.4336 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.5664, | |
| "fpr": 0.4336 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.5664, | |
| "fpr": 0.4336 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.5664, | |
| "fpr": 0.4336 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9345, | |
| "fpr": 0.0655 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9345, | |
| "fpr": 0.0655 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "fpr": 0.0652 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "fpr": 0.0652 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9412, | |
| "recall": 0.8158, | |
| "fpr": 0.0102, | |
| "auroc": 0.9901, | |
| "pauc10": 0.92, | |
| "precision": 0.9688, | |
| "f1": 0.8857 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9412, | |
| "recall": 0.8158, | |
| "fpr": 0.0102, | |
| "auroc": 0.9901, | |
| "pauc10": 0.92, | |
| "precision": 0.9688, | |
| "f1": 0.8857 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9084, | |
| "recall": 0.8417, | |
| "fpr": 0.0163, | |
| "auroc": 0.9791, | |
| "pauc10": 0.8529, | |
| "precision": 0.9832, | |
| "f1": 0.907 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9084, | |
| "recall": 0.8417, | |
| "fpr": 0.0163, | |
| "auroc": 0.9791, | |
| "pauc10": 0.8529, | |
| "precision": 0.9832, | |
| "f1": 0.907 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6724, | |
| "recall": 0.3667, | |
| "fpr": 0.0, | |
| "auroc": 0.9015, | |
| "pauc10": 0.7149, | |
| "precision": 1.0, | |
| "f1": 0.5366 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6724, | |
| "recall": 0.3667, | |
| "fpr": 0.0, | |
| "auroc": 0.9015, | |
| "pauc10": 0.7149, | |
| "precision": 1.0, | |
| "f1": 0.5366 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6724, | |
| "recall": 0.3667, | |
| "fpr": 0.0, | |
| "auroc": 0.9015, | |
| "pauc10": 0.7149, | |
| "precision": 1.0, | |
| "f1": 0.5366 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6724, | |
| "recall": 0.3667, | |
| "fpr": 0.0, | |
| "auroc": 0.9015, | |
| "pauc10": 0.7149, | |
| "precision": 1.0, | |
| "f1": 0.5366 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.5211, | |
| "recall": 0.5211 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.5211, | |
| "recall": 0.5211 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.5383, | |
| "recall": 0.5383 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.5383, | |
| "recall": 0.5383 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.6333, | |
| "recall": 0.8417, | |
| "fpr": 0.575, | |
| "auroc": 0.7238, | |
| "pauc10": 0.2965, | |
| "precision": 0.5941, | |
| "f1": 0.6966 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.6333, | |
| "recall": 0.8417, | |
| "fpr": 0.575, | |
| "auroc": 0.7238, | |
| "pauc10": 0.2965, | |
| "precision": 0.5941, | |
| "f1": 0.6966 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.6333, | |
| "recall": 0.8417, | |
| "fpr": 0.575, | |
| "auroc": 0.7238, | |
| "pauc10": 0.2965, | |
| "precision": 0.5941, | |
| "f1": 0.6966 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.6333, | |
| "recall": 0.8417, | |
| "fpr": 0.575, | |
| "auroc": 0.7238, | |
| "pauc10": 0.2965, | |
| "precision": 0.5941, | |
| "f1": 0.6966 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2036, | |
| "recall": 0.1667, | |
| "fpr": 0.1742, | |
| "auroc": 0.4219, | |
| "pauc10": 0.0428, | |
| "precision": 0.9416, | |
| "f1": 0.2832 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2036, | |
| "recall": 0.1667, | |
| "fpr": 0.1742, | |
| "auroc": 0.4219, | |
| "pauc10": 0.0428, | |
| "precision": 0.9416, | |
| "f1": 0.2832 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2036, | |
| "recall": 0.1667, | |
| "fpr": 0.1742, | |
| "auroc": 0.4219, | |
| "pauc10": 0.0428, | |
| "precision": 0.9416, | |
| "f1": 0.2832 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2036, | |
| "recall": 0.1667, | |
| "fpr": 0.1742, | |
| "auroc": 0.4219, | |
| "pauc10": 0.0428, | |
| "precision": 0.9416, | |
| "f1": 0.2832 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.598, | |
| "recall": 0.2053, | |
| "fpr": 0.0093, | |
| "auroc": 0.7224, | |
| "pauc10": 0.2885, | |
| "precision": 0.9565, | |
| "f1": 0.3381 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.598, | |
| "recall": 0.2053, | |
| "fpr": 0.0093, | |
| "auroc": 0.7224, | |
| "pauc10": 0.2885, | |
| "precision": 0.9565, | |
| "f1": 0.3381 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.598, | |
| "recall": 0.2053, | |
| "fpr": 0.0093, | |
| "auroc": 0.7224, | |
| "pauc10": 0.2885, | |
| "precision": 0.9565, | |
| "f1": 0.3381 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.598, | |
| "recall": 0.2053, | |
| "fpr": 0.0093, | |
| "auroc": 0.7224, | |
| "pauc10": 0.2885, | |
| "precision": 0.9565, | |
| "f1": 0.3381 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.456, | |
| "recall": 0.456 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.456, | |
| "recall": 0.456 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.456, | |
| "recall": 0.456 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.456, | |
| "recall": 0.456 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9085, | |
| "recall": 0.8636, | |
| "fpr": 0.0282, | |
| "auroc": 0.9764, | |
| "pauc10": 0.8907, | |
| "precision": 0.9774, | |
| "f1": 0.917 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9085, | |
| "recall": 0.8636, | |
| "fpr": 0.0282, | |
| "auroc": 0.9764, | |
| "pauc10": 0.8907, | |
| "precision": 0.9774, | |
| "f1": 0.917 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9087, | |
| "recall": 0.8641, | |
| "fpr": 0.0282, | |
| "auroc": 0.9764, | |
| "pauc10": 0.8909, | |
| "precision": 0.9775, | |
| "f1": 0.9173 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9087, | |
| "recall": 0.8641, | |
| "fpr": 0.0282, | |
| "auroc": 0.9764, | |
| "pauc10": 0.8909, | |
| "precision": 0.9775, | |
| "f1": 0.9173 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8515, | |
| "recall": 0.7069, | |
| "fpr": 0.0038, | |
| "auroc": 0.9776, | |
| "pauc10": 0.893, | |
| "precision": 0.9946, | |
| "f1": 0.8265 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8515, | |
| "recall": 0.7069, | |
| "fpr": 0.0038, | |
| "auroc": 0.9776, | |
| "pauc10": 0.893, | |
| "precision": 0.9946, | |
| "f1": 0.8265 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8515, | |
| "recall": 0.7069, | |
| "fpr": 0.0038, | |
| "auroc": 0.9776, | |
| "pauc10": 0.893, | |
| "precision": 0.9946, | |
| "f1": 0.8265 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8515, | |
| "recall": 0.7069, | |
| "fpr": 0.0038, | |
| "auroc": 0.9776, | |
| "pauc10": 0.893, | |
| "precision": 0.9946, | |
| "f1": 0.8265 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7266, | |
| "recall": 0.9107, | |
| "fpr": 0.4679, | |
| "auroc": 0.7579, | |
| "pauc10": 0.0626, | |
| "precision": 0.6728, | |
| "f1": 0.7739 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7266, | |
| "recall": 0.9107, | |
| "fpr": 0.4679, | |
| "auroc": 0.7579, | |
| "pauc10": 0.0626, | |
| "precision": 0.6728, | |
| "f1": 0.7739 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7266, | |
| "recall": 0.9107, | |
| "fpr": 0.4679, | |
| "auroc": 0.7579, | |
| "pauc10": 0.0626, | |
| "precision": 0.6728, | |
| "f1": 0.7739 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7266, | |
| "recall": 0.9107, | |
| "fpr": 0.4679, | |
| "auroc": 0.7579, | |
| "pauc10": 0.0626, | |
| "precision": 0.6728, | |
| "f1": 0.7739 | |
| } | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.5664, | |
| "fpr": 0.4336 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.5664, | |
| "fpr": 0.4336 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.5664, | |
| "fpr": 0.4336 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.5664, | |
| "fpr": 0.4336 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9345, | |
| "fpr": 0.0655 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9345, | |
| "fpr": 0.0655 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "fpr": 0.0652 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9348, | |
| "fpr": 0.0652 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9412, | |
| "recall": 0.8158, | |
| "fpr": 0.0102, | |
| "auroc": 0.9933, | |
| "pauc10": 0.9329, | |
| "precision": 0.9688, | |
| "f1": 0.8857 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9412, | |
| "recall": 0.8158, | |
| "fpr": 0.0102, | |
| "auroc": 0.9933, | |
| "pauc10": 0.9329, | |
| "precision": 0.9688, | |
| "f1": 0.8857 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9122, | |
| "recall": 0.8489, | |
| "fpr": 0.0163, | |
| "auroc": 0.98, | |
| "pauc10": 0.8554, | |
| "precision": 0.9833, | |
| "f1": 0.9112 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9122, | |
| "recall": 0.8489, | |
| "fpr": 0.0163, | |
| "auroc": 0.98, | |
| "pauc10": 0.8554, | |
| "precision": 0.9833, | |
| "f1": 0.9112 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6724, | |
| "recall": 0.3667, | |
| "fpr": 0.0, | |
| "auroc": 0.9015, | |
| "pauc10": 0.7149, | |
| "precision": 1.0, | |
| "f1": 0.5366 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6724, | |
| "recall": 0.3667, | |
| "fpr": 0.0, | |
| "auroc": 0.9015, | |
| "pauc10": 0.7149, | |
| "precision": 1.0, | |
| "f1": 0.5366 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6724, | |
| "recall": 0.3667, | |
| "fpr": 0.0, | |
| "auroc": 0.9015, | |
| "pauc10": 0.7149, | |
| "precision": 1.0, | |
| "f1": 0.5366 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6724, | |
| "recall": 0.3667, | |
| "fpr": 0.0, | |
| "auroc": 0.9015, | |
| "pauc10": 0.7149, | |
| "precision": 1.0, | |
| "f1": 0.5366 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.5229, | |
| "recall": 0.5229 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.5229, | |
| "recall": 0.5229 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.5404, | |
| "recall": 0.5404 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.5404, | |
| "recall": 0.5404 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.6333, | |
| "recall": 0.8417, | |
| "fpr": 0.575, | |
| "auroc": 0.7209, | |
| "pauc10": 0.2729, | |
| "precision": 0.5941, | |
| "f1": 0.6966 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.6333, | |
| "recall": 0.8417, | |
| "fpr": 0.575, | |
| "auroc": 0.7209, | |
| "pauc10": 0.2729, | |
| "precision": 0.5941, | |
| "f1": 0.6966 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.6333, | |
| "recall": 0.8417, | |
| "fpr": 0.575, | |
| "auroc": 0.7209, | |
| "pauc10": 0.2729, | |
| "precision": 0.5941, | |
| "f1": 0.6966 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.6333, | |
| "recall": 0.8417, | |
| "fpr": 0.575, | |
| "auroc": 0.7209, | |
| "pauc10": 0.2729, | |
| "precision": 0.5941, | |
| "f1": 0.6966 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2215, | |
| "recall": 0.187, | |
| "fpr": 0.1966, | |
| "auroc": 0.4386, | |
| "pauc10": 0.0428, | |
| "precision": 0.9413, | |
| "f1": 0.312 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2215, | |
| "recall": 0.187, | |
| "fpr": 0.1966, | |
| "auroc": 0.4386, | |
| "pauc10": 0.0428, | |
| "precision": 0.9413, | |
| "f1": 0.312 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2215, | |
| "recall": 0.187, | |
| "fpr": 0.1966, | |
| "auroc": 0.4386, | |
| "pauc10": 0.0428, | |
| "precision": 0.9413, | |
| "f1": 0.312 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2215, | |
| "recall": 0.187, | |
| "fpr": 0.1966, | |
| "auroc": 0.4386, | |
| "pauc10": 0.0428, | |
| "precision": 0.9413, | |
| "f1": 0.312 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.608, | |
| "recall": 0.2267, | |
| "fpr": 0.0107, | |
| "auroc": 0.7321, | |
| "pauc10": 0.301, | |
| "precision": 0.9551, | |
| "f1": 0.3664 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.608, | |
| "recall": 0.2267, | |
| "fpr": 0.0107, | |
| "auroc": 0.7321, | |
| "pauc10": 0.301, | |
| "precision": 0.9551, | |
| "f1": 0.3664 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.608, | |
| "recall": 0.2267, | |
| "fpr": 0.0107, | |
| "auroc": 0.7321, | |
| "pauc10": 0.301, | |
| "precision": 0.9551, | |
| "f1": 0.3664 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.608, | |
| "recall": 0.2267, | |
| "fpr": 0.0107, | |
| "auroc": 0.7321, | |
| "pauc10": 0.301, | |
| "precision": 0.9551, | |
| "f1": 0.3664 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.485, | |
| "recall": 0.485 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.485, | |
| "recall": 0.485 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.485, | |
| "recall": 0.485 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.485, | |
| "recall": 0.485 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9085, | |
| "recall": 0.8636, | |
| "fpr": 0.0282, | |
| "auroc": 0.9764, | |
| "pauc10": 0.8908, | |
| "precision": 0.9774, | |
| "f1": 0.917 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9085, | |
| "recall": 0.8636, | |
| "fpr": 0.0282, | |
| "auroc": 0.9764, | |
| "pauc10": 0.8908, | |
| "precision": 0.9774, | |
| "f1": 0.917 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9087, | |
| "recall": 0.8641, | |
| "fpr": 0.0282, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8912, | |
| "precision": 0.9775, | |
| "f1": 0.9173 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9087, | |
| "recall": 0.8641, | |
| "fpr": 0.0282, | |
| "auroc": 0.9765, | |
| "pauc10": 0.8912, | |
| "precision": 0.9775, | |
| "f1": 0.9173 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8515, | |
| "recall": 0.7069, | |
| "fpr": 0.0038, | |
| "auroc": 0.9776, | |
| "pauc10": 0.893, | |
| "precision": 0.9946, | |
| "f1": 0.8265 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8515, | |
| "recall": 0.7069, | |
| "fpr": 0.0038, | |
| "auroc": 0.9776, | |
| "pauc10": 0.893, | |
| "precision": 0.9946, | |
| "f1": 0.8265 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8515, | |
| "recall": 0.7069, | |
| "fpr": 0.0038, | |
| "auroc": 0.9776, | |
| "pauc10": 0.893, | |
| "precision": 0.9946, | |
| "f1": 0.8265 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8515, | |
| "recall": 0.7069, | |
| "fpr": 0.0038, | |
| "auroc": 0.9776, | |
| "pauc10": 0.893, | |
| "precision": 0.9946, | |
| "f1": 0.8265 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7266, | |
| "recall": 0.9107, | |
| "fpr": 0.4679, | |
| "auroc": 0.7579, | |
| "pauc10": 0.0626, | |
| "precision": 0.6728, | |
| "f1": 0.7739 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7266, | |
| "recall": 0.9107, | |
| "fpr": 0.4679, | |
| "auroc": 0.7579, | |
| "pauc10": 0.0626, | |
| "precision": 0.6728, | |
| "f1": 0.7739 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7266, | |
| "recall": 0.9107, | |
| "fpr": 0.4679, | |
| "auroc": 0.7579, | |
| "pauc10": 0.0626, | |
| "precision": 0.6728, | |
| "f1": 0.7739 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7266, | |
| "recall": 0.9107, | |
| "fpr": 0.4679, | |
| "auroc": 0.7579, | |
| "pauc10": 0.0626, | |
| "precision": 0.6728, | |
| "f1": 0.7739 | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "deepset": { | |
| "serve": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2861, | |
| "fpr": 0.7139 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2861, | |
| "fpr": 0.7139 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2861, | |
| "fpr": 0.7139 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2861, | |
| "fpr": 0.7139 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8026, | |
| "fpr": 0.1974 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8026, | |
| "fpr": 0.1974 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8029, | |
| "fpr": 0.1971 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8029, | |
| "fpr": 0.1971 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.3235, | |
| "recall": 1.0, | |
| "fpr": 0.9388, | |
| "auroc": 0.747, | |
| "pauc10": 0.0274, | |
| "precision": 0.2923, | |
| "f1": 0.4524 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.3235, | |
| "recall": 1.0, | |
| "fpr": 0.9388, | |
| "auroc": 0.747, | |
| "pauc10": 0.0274, | |
| "precision": 0.2923, | |
| "f1": 0.4524 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.5573, | |
| "recall": 1.0, | |
| "fpr": 0.9431, | |
| "auroc": 0.7886, | |
| "pauc10": 0.0779, | |
| "precision": 0.5451, | |
| "f1": 0.7056 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.5573, | |
| "recall": 1.0, | |
| "fpr": 0.9431, | |
| "auroc": 0.7886, | |
| "pauc10": 0.0779, | |
| "precision": 0.5451, | |
| "f1": 0.7056 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.9914, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 0.997, | |
| "pauc10": 0.9833, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.9914, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 0.997, | |
| "pauc10": 0.9833, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.9914, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 0.997, | |
| "pauc10": 0.9833, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.9914, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 0.997, | |
| "pauc10": 0.9833, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.9961, | |
| "recall": 0.9961 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.9961, | |
| "recall": 0.9961 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.9959, | |
| "recall": 0.9959 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.9959, | |
| "recall": 0.9959 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5029, | |
| "pauc10": 0.0403, | |
| "precision": 0.5, | |
| "f1": 0.6667 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5029, | |
| "pauc10": 0.0403, | |
| "precision": 0.5, | |
| "f1": 0.6667 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5029, | |
| "pauc10": 0.0403, | |
| "precision": 0.5, | |
| "f1": 0.6667 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5029, | |
| "pauc10": 0.0403, | |
| "precision": 0.5, | |
| "f1": 0.6667 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5771, | |
| "pauc10": 0.1614, | |
| "precision": 0.944, | |
| "f1": 0.9712 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5771, | |
| "pauc10": 0.1614, | |
| "precision": 0.944, | |
| "f1": 0.9712 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5771, | |
| "pauc10": 0.1614, | |
| "precision": 0.944, | |
| "f1": 0.9712 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5771, | |
| "pauc10": 0.1614, | |
| "precision": 0.944, | |
| "f1": 0.9712 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5113, | |
| "recall": 1.0, | |
| "fpr": 0.9773, | |
| "auroc": 0.836, | |
| "pauc10": 0.3307, | |
| "precision": 0.5057, | |
| "f1": 0.6717 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5113, | |
| "recall": 1.0, | |
| "fpr": 0.9773, | |
| "auroc": 0.836, | |
| "pauc10": 0.3307, | |
| "precision": 0.5057, | |
| "f1": 0.6717 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5113, | |
| "recall": 1.0, | |
| "fpr": 0.9773, | |
| "auroc": 0.836, | |
| "pauc10": 0.3307, | |
| "precision": 0.5057, | |
| "f1": 0.6717 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5113, | |
| "recall": 1.0, | |
| "fpr": 0.9773, | |
| "auroc": 0.836, | |
| "pauc10": 0.3307, | |
| "precision": 0.5057, | |
| "f1": 0.6717 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "recall": 1.0 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "recall": 1.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "recall": 1.0 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "recall": 1.0 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.8202, | |
| "recall": 0.9818, | |
| "fpr": 0.4077, | |
| "auroc": 0.9581, | |
| "pauc10": 0.7549, | |
| "precision": 0.7725, | |
| "f1": 0.8647 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.8202, | |
| "recall": 0.9818, | |
| "fpr": 0.4077, | |
| "auroc": 0.9581, | |
| "pauc10": 0.7549, | |
| "precision": 0.7725, | |
| "f1": 0.8647 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.8206, | |
| "recall": 0.9819, | |
| "fpr": 0.4077, | |
| "auroc": 0.9582, | |
| "pauc10": 0.755, | |
| "precision": 0.7732, | |
| "f1": 0.8651 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.8206, | |
| "recall": 0.9819, | |
| "fpr": 0.4077, | |
| "auroc": 0.9582, | |
| "pauc10": 0.755, | |
| "precision": 0.7732, | |
| "f1": 0.8651 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8872, | |
| "recall": 0.8021, | |
| "fpr": 0.0276, | |
| "auroc": 0.9609, | |
| "pauc10": 0.8162, | |
| "precision": 0.9667, | |
| "f1": 0.8768 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8872, | |
| "recall": 0.8021, | |
| "fpr": 0.0276, | |
| "auroc": 0.9609, | |
| "pauc10": 0.8162, | |
| "precision": 0.9667, | |
| "f1": 0.8768 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8872, | |
| "recall": 0.8021, | |
| "fpr": 0.0276, | |
| "auroc": 0.9609, | |
| "pauc10": 0.8162, | |
| "precision": 0.9667, | |
| "f1": 0.8768 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8872, | |
| "recall": 0.8021, | |
| "fpr": 0.0276, | |
| "auroc": 0.9609, | |
| "pauc10": 0.8162, | |
| "precision": 0.9667, | |
| "f1": 0.8768 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.5358, | |
| "recall": 1.0, | |
| "fpr": 0.9547, | |
| "auroc": 0.923, | |
| "pauc10": 0.5208, | |
| "precision": 0.5253, | |
| "f1": 0.6888 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.5358, | |
| "recall": 1.0, | |
| "fpr": 0.9547, | |
| "auroc": 0.923, | |
| "pauc10": 0.5208, | |
| "precision": 0.5253, | |
| "f1": 0.6888 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.5358, | |
| "recall": 1.0, | |
| "fpr": 0.9547, | |
| "auroc": 0.923, | |
| "pauc10": 0.5208, | |
| "precision": 0.5253, | |
| "f1": 0.6888 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.5358, | |
| "recall": 1.0, | |
| "fpr": 0.9547, | |
| "auroc": 0.923, | |
| "pauc10": 0.5208, | |
| "precision": 0.5253, | |
| "f1": 0.6888 | |
| } | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2861, | |
| "fpr": 0.7139 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2861, | |
| "fpr": 0.7139 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2861, | |
| "fpr": 0.7139 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.2861, | |
| "fpr": 0.7139 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "fpr": 0.0 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8026, | |
| "fpr": 0.1974 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8026, | |
| "fpr": 0.1974 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8029, | |
| "fpr": 0.1971 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8029, | |
| "fpr": 0.1971 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.3235, | |
| "recall": 1.0, | |
| "fpr": 0.9388, | |
| "auroc": 0.8201, | |
| "pauc10": 0.0677, | |
| "precision": 0.2923, | |
| "f1": 0.4524 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.3235, | |
| "recall": 1.0, | |
| "fpr": 0.9388, | |
| "auroc": 0.8201, | |
| "pauc10": 0.0677, | |
| "precision": 0.2923, | |
| "f1": 0.4524 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.5573, | |
| "recall": 1.0, | |
| "fpr": 0.9431, | |
| "auroc": 0.8436, | |
| "pauc10": 0.1288, | |
| "precision": 0.5451, | |
| "f1": 0.7056 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.5573, | |
| "recall": 1.0, | |
| "fpr": 0.9431, | |
| "auroc": 0.8436, | |
| "pauc10": 0.1288, | |
| "precision": 0.5451, | |
| "f1": 0.7056 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.9914, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 0.997, | |
| "pauc10": 0.9833, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.9914, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 0.997, | |
| "pauc10": 0.9833, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.9914, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 0.997, | |
| "pauc10": 0.9833, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.9914, | |
| "recall": 0.9833, | |
| "fpr": 0.0, | |
| "auroc": 0.997, | |
| "pauc10": 0.9833, | |
| "precision": 1.0, | |
| "f1": 0.9916 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.9961, | |
| "recall": 0.9961 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.9961, | |
| "recall": 0.9961 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.9959, | |
| "recall": 0.9959 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.9959, | |
| "recall": 0.9959 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.4989, | |
| "pauc10": 0.0403, | |
| "precision": 0.5, | |
| "f1": 0.6667 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.4989, | |
| "pauc10": 0.0403, | |
| "precision": 0.5, | |
| "f1": 0.6667 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.4989, | |
| "pauc10": 0.0403, | |
| "precision": 0.5, | |
| "f1": 0.6667 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.5, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.4989, | |
| "pauc10": 0.0403, | |
| "precision": 0.5, | |
| "f1": 0.6667 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5743, | |
| "pauc10": 0.163, | |
| "precision": 0.944, | |
| "f1": 0.9712 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5743, | |
| "pauc10": 0.163, | |
| "precision": 0.944, | |
| "f1": 0.9712 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5743, | |
| "pauc10": 0.163, | |
| "precision": 0.944, | |
| "f1": 0.9712 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "recall": 1.0, | |
| "fpr": 1.0, | |
| "auroc": 0.5743, | |
| "pauc10": 0.163, | |
| "precision": 0.944, | |
| "f1": 0.9712 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5113, | |
| "recall": 1.0, | |
| "fpr": 0.9773, | |
| "auroc": 0.82, | |
| "pauc10": 0.2662, | |
| "precision": 0.5057, | |
| "f1": 0.6717 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5113, | |
| "recall": 1.0, | |
| "fpr": 0.9773, | |
| "auroc": 0.82, | |
| "pauc10": 0.2662, | |
| "precision": 0.5057, | |
| "f1": 0.6717 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5113, | |
| "recall": 1.0, | |
| "fpr": 0.9773, | |
| "auroc": 0.82, | |
| "pauc10": 0.2662, | |
| "precision": 0.5057, | |
| "f1": 0.6717 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5113, | |
| "recall": 1.0, | |
| "fpr": 0.9773, | |
| "auroc": 0.82, | |
| "pauc10": 0.2662, | |
| "precision": 0.5057, | |
| "f1": 0.6717 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "recall": 1.0 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "recall": 1.0 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "recall": 1.0 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 1.0, | |
| "recall": 1.0 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.8202, | |
| "recall": 0.9818, | |
| "fpr": 0.4077, | |
| "auroc": 0.9582, | |
| "pauc10": 0.7561, | |
| "precision": 0.7725, | |
| "f1": 0.8647 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.8202, | |
| "recall": 0.9818, | |
| "fpr": 0.4077, | |
| "auroc": 0.9582, | |
| "pauc10": 0.7561, | |
| "precision": 0.7725, | |
| "f1": 0.8647 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.8206, | |
| "recall": 0.9819, | |
| "fpr": 0.4077, | |
| "auroc": 0.9583, | |
| "pauc10": 0.7565, | |
| "precision": 0.7732, | |
| "f1": 0.8651 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.8206, | |
| "recall": 0.9819, | |
| "fpr": 0.4077, | |
| "auroc": 0.9583, | |
| "pauc10": 0.7565, | |
| "precision": 0.7732, | |
| "f1": 0.8651 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8872, | |
| "recall": 0.8021, | |
| "fpr": 0.0276, | |
| "auroc": 0.9609, | |
| "pauc10": 0.8162, | |
| "precision": 0.9667, | |
| "f1": 0.8768 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8872, | |
| "recall": 0.8021, | |
| "fpr": 0.0276, | |
| "auroc": 0.9609, | |
| "pauc10": 0.8162, | |
| "precision": 0.9667, | |
| "f1": 0.8768 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8872, | |
| "recall": 0.8021, | |
| "fpr": 0.0276, | |
| "auroc": 0.9609, | |
| "pauc10": 0.8162, | |
| "precision": 0.9667, | |
| "f1": 0.8768 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8872, | |
| "recall": 0.8021, | |
| "fpr": 0.0276, | |
| "auroc": 0.9609, | |
| "pauc10": 0.8162, | |
| "precision": 0.9667, | |
| "f1": 0.8768 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.5358, | |
| "recall": 1.0, | |
| "fpr": 0.9547, | |
| "auroc": 0.923, | |
| "pauc10": 0.5208, | |
| "precision": 0.5253, | |
| "f1": 0.6888 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.5358, | |
| "recall": 1.0, | |
| "fpr": 0.9547, | |
| "auroc": 0.923, | |
| "pauc10": 0.5208, | |
| "precision": 0.5253, | |
| "f1": 0.6888 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.5358, | |
| "recall": 1.0, | |
| "fpr": 0.9547, | |
| "auroc": 0.923, | |
| "pauc10": 0.5208, | |
| "precision": 0.5253, | |
| "f1": 0.6888 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.5358, | |
| "recall": 1.0, | |
| "fpr": 0.9547, | |
| "auroc": 0.923, | |
| "pauc10": 0.5208, | |
| "precision": 0.5253, | |
| "f1": 0.6888 | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "wolf-defender": { | |
| "serve": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9233, | |
| "fpr": 0.0767 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9233, | |
| "fpr": 0.0767 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9233, | |
| "fpr": 0.0767 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9233, | |
| "fpr": 0.0767 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9637, | |
| "fpr": 0.0363 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9637, | |
| "fpr": 0.0363 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9644, | |
| "fpr": 0.0356 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9644, | |
| "fpr": 0.0356 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8559, | |
| "fpr": 0.1441 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8559, | |
| "fpr": 0.1441 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "fpr": 0.1433 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "fpr": 0.1433 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9412, | |
| "recall": 0.8947, | |
| "fpr": 0.0408, | |
| "auroc": 0.9608, | |
| "pauc10": 0.8142, | |
| "precision": 0.8947, | |
| "f1": 0.8947 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9412, | |
| "recall": 0.8947, | |
| "fpr": 0.0408, | |
| "auroc": 0.9608, | |
| "pauc10": 0.8142, | |
| "precision": 0.8947, | |
| "f1": 0.8947 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9389, | |
| "recall": 0.9353, | |
| "fpr": 0.0569, | |
| "auroc": 0.9744, | |
| "pauc10": 0.8576, | |
| "precision": 0.9489, | |
| "f1": 0.942 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9389, | |
| "recall": 0.9353, | |
| "fpr": 0.0569, | |
| "auroc": 0.9744, | |
| "pauc10": 0.8576, | |
| "precision": 0.9489, | |
| "f1": 0.942 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.7931, | |
| "recall": 0.6, | |
| "fpr": 0.0, | |
| "auroc": 0.9685, | |
| "pauc10": 0.7869, | |
| "precision": 1.0, | |
| "f1": 0.75 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.7931, | |
| "recall": 0.6, | |
| "fpr": 0.0, | |
| "auroc": 0.9685, | |
| "pauc10": 0.7869, | |
| "precision": 1.0, | |
| "f1": 0.75 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.7931, | |
| "recall": 0.6, | |
| "fpr": 0.0, | |
| "auroc": 0.9685, | |
| "pauc10": 0.7869, | |
| "precision": 1.0, | |
| "f1": 0.75 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.7931, | |
| "recall": 0.6, | |
| "fpr": 0.0, | |
| "auroc": 0.9685, | |
| "pauc10": 0.7869, | |
| "precision": 1.0, | |
| "f1": 0.75 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.834, | |
| "recall": 0.834 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.834, | |
| "recall": 0.834 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8409, | |
| "recall": 0.8409 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8409, | |
| "recall": 0.8409 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.775, | |
| "recall": 0.8333, | |
| "fpr": 0.2833, | |
| "auroc": 0.8766, | |
| "pauc10": 0.5632, | |
| "precision": 0.7463, | |
| "f1": 0.7874 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.775, | |
| "recall": 0.8333, | |
| "fpr": 0.2833, | |
| "auroc": 0.8766, | |
| "pauc10": 0.5632, | |
| "precision": 0.7463, | |
| "f1": 0.7874 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.775, | |
| "recall": 0.8333, | |
| "fpr": 0.2833, | |
| "auroc": 0.8766, | |
| "pauc10": 0.5632, | |
| "precision": 0.7463, | |
| "f1": 0.7874 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.775, | |
| "recall": 0.8333, | |
| "fpr": 0.2833, | |
| "auroc": 0.8766, | |
| "pauc10": 0.5632, | |
| "precision": 0.7463, | |
| "f1": 0.7874 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2766, | |
| "recall": 0.2397, | |
| "fpr": 0.1011, | |
| "auroc": 0.7124, | |
| "pauc10": 0.1747, | |
| "precision": 0.9756, | |
| "f1": 0.3848 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2766, | |
| "recall": 0.2397, | |
| "fpr": 0.1011, | |
| "auroc": 0.7124, | |
| "pauc10": 0.1747, | |
| "precision": 0.9756, | |
| "f1": 0.3848 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2766, | |
| "recall": 0.2397, | |
| "fpr": 0.1011, | |
| "auroc": 0.7124, | |
| "pauc10": 0.1747, | |
| "precision": 0.9756, | |
| "f1": 0.3848 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2766, | |
| "recall": 0.2397, | |
| "fpr": 0.1011, | |
| "auroc": 0.7124, | |
| "pauc10": 0.1747, | |
| "precision": 0.9756, | |
| "f1": 0.3848 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.6093, | |
| "recall": 0.22, | |
| "fpr": 0.0013, | |
| "auroc": 0.8604, | |
| "pauc10": 0.5737, | |
| "precision": 0.994, | |
| "f1": 0.3603 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.6093, | |
| "recall": 0.22, | |
| "fpr": 0.0013, | |
| "auroc": 0.8604, | |
| "pauc10": 0.5737, | |
| "precision": 0.994, | |
| "f1": 0.3603 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.6093, | |
| "recall": 0.22, | |
| "fpr": 0.0013, | |
| "auroc": 0.8604, | |
| "pauc10": 0.5737, | |
| "precision": 0.994, | |
| "f1": 0.3603 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.6093, | |
| "recall": 0.22, | |
| "fpr": 0.0013, | |
| "auroc": 0.8604, | |
| "pauc10": 0.5737, | |
| "precision": 0.994, | |
| "f1": 0.3603 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.898, | |
| "recall": 0.898 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.898, | |
| "recall": 0.898 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.898, | |
| "recall": 0.898 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.898, | |
| "recall": 0.898 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9287, | |
| "recall": 0.9073, | |
| "fpr": 0.041, | |
| "auroc": 0.9782, | |
| "pauc10": 0.8777, | |
| "precision": 0.9689, | |
| "f1": 0.9371 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9287, | |
| "recall": 0.9073, | |
| "fpr": 0.041, | |
| "auroc": 0.9782, | |
| "pauc10": 0.8777, | |
| "precision": 0.9689, | |
| "f1": 0.9371 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9289, | |
| "recall": 0.9076, | |
| "fpr": 0.041, | |
| "auroc": 0.9782, | |
| "pauc10": 0.8782, | |
| "precision": 0.9691, | |
| "f1": 0.9373 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9289, | |
| "recall": 0.9076, | |
| "fpr": 0.041, | |
| "auroc": 0.9782, | |
| "pauc10": 0.8782, | |
| "precision": 0.9691, | |
| "f1": 0.9373 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8953, | |
| "recall": 0.7926, | |
| "fpr": 0.0019, | |
| "auroc": 0.993, | |
| "pauc10": 0.9552, | |
| "precision": 0.9976, | |
| "f1": 0.8834 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8953, | |
| "recall": 0.7926, | |
| "fpr": 0.0019, | |
| "auroc": 0.993, | |
| "pauc10": 0.9552, | |
| "precision": 0.9976, | |
| "f1": 0.8834 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8953, | |
| "recall": 0.7926, | |
| "fpr": 0.0019, | |
| "auroc": 0.993, | |
| "pauc10": 0.9552, | |
| "precision": 0.9976, | |
| "f1": 0.8834 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8953, | |
| "recall": 0.7926, | |
| "fpr": 0.0019, | |
| "auroc": 0.993, | |
| "pauc10": 0.9552, | |
| "precision": 0.9976, | |
| "f1": 0.8834 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7963, | |
| "recall": 0.7929, | |
| "fpr": 0.2, | |
| "auroc": 0.8544, | |
| "pauc10": 0.5338, | |
| "precision": 0.8073, | |
| "f1": 0.8 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7963, | |
| "recall": 0.7929, | |
| "fpr": 0.2, | |
| "auroc": 0.8544, | |
| "pauc10": 0.5338, | |
| "precision": 0.8073, | |
| "f1": 0.8 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7963, | |
| "recall": 0.7929, | |
| "fpr": 0.2, | |
| "auroc": 0.8544, | |
| "pauc10": 0.5338, | |
| "precision": 0.8073, | |
| "f1": 0.8 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7963, | |
| "recall": 0.7929, | |
| "fpr": 0.2, | |
| "auroc": 0.8544, | |
| "pauc10": 0.5338, | |
| "precision": 0.8073, | |
| "f1": 0.8 | |
| } | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9233, | |
| "fpr": 0.0767 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9233, | |
| "fpr": 0.0767 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9233, | |
| "fpr": 0.0767 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9233, | |
| "fpr": 0.0767 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9637, | |
| "fpr": 0.0363 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9637, | |
| "fpr": 0.0363 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9644, | |
| "fpr": 0.0356 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.9644, | |
| "fpr": 0.0356 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8559, | |
| "fpr": 0.1441 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8559, | |
| "fpr": 0.1441 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "fpr": 0.1433 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "fpr": 0.1433 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9485, | |
| "recall": 0.8947, | |
| "fpr": 0.0306, | |
| "auroc": 0.9637, | |
| "pauc10": 0.8292, | |
| "precision": 0.9189, | |
| "f1": 0.9067 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9485, | |
| "recall": 0.8947, | |
| "fpr": 0.0306, | |
| "auroc": 0.9637, | |
| "pauc10": 0.8292, | |
| "precision": 0.9189, | |
| "f1": 0.9067 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9427, | |
| "recall": 0.9353, | |
| "fpr": 0.0488, | |
| "auroc": 0.9765, | |
| "pauc10": 0.856, | |
| "precision": 0.9559, | |
| "f1": 0.9455 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.9427, | |
| "recall": 0.9353, | |
| "fpr": 0.0488, | |
| "auroc": 0.9765, | |
| "pauc10": 0.856, | |
| "precision": 0.9559, | |
| "f1": 0.9455 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.7931, | |
| "recall": 0.6, | |
| "fpr": 0.0, | |
| "auroc": 0.964, | |
| "pauc10": 0.7869, | |
| "precision": 1.0, | |
| "f1": 0.75 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.7931, | |
| "recall": 0.6, | |
| "fpr": 0.0, | |
| "auroc": 0.964, | |
| "pauc10": 0.7869, | |
| "precision": 1.0, | |
| "f1": 0.75 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.7931, | |
| "recall": 0.6, | |
| "fpr": 0.0, | |
| "auroc": 0.964, | |
| "pauc10": 0.7869, | |
| "precision": 1.0, | |
| "f1": 0.75 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.7931, | |
| "recall": 0.6, | |
| "fpr": 0.0, | |
| "auroc": 0.964, | |
| "pauc10": 0.7869, | |
| "precision": 1.0, | |
| "f1": 0.75 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.8301, | |
| "recall": 0.8301 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.8301, | |
| "recall": 0.8301 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8368, | |
| "recall": 0.8368 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.8368, | |
| "recall": 0.8368 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.775, | |
| "recall": 0.8333, | |
| "fpr": 0.2833, | |
| "auroc": 0.8791, | |
| "pauc10": 0.5882, | |
| "precision": 0.7463, | |
| "f1": 0.7874 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.775, | |
| "recall": 0.8333, | |
| "fpr": 0.2833, | |
| "auroc": 0.8791, | |
| "pauc10": 0.5882, | |
| "precision": 0.7463, | |
| "f1": 0.7874 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.775, | |
| "recall": 0.8333, | |
| "fpr": 0.2833, | |
| "auroc": 0.8791, | |
| "pauc10": 0.5882, | |
| "precision": 0.7463, | |
| "f1": 0.7874 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.775, | |
| "recall": 0.8333, | |
| "fpr": 0.2833, | |
| "auroc": 0.8791, | |
| "pauc10": 0.5882, | |
| "precision": 0.7463, | |
| "f1": 0.7874 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2001, | |
| "recall": 0.1567, | |
| "fpr": 0.0674, | |
| "auroc": 0.6246, | |
| "pauc10": 0.1339, | |
| "precision": 0.9751, | |
| "f1": 0.27 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2001, | |
| "recall": 0.1567, | |
| "fpr": 0.0674, | |
| "auroc": 0.6246, | |
| "pauc10": 0.1339, | |
| "precision": 0.9751, | |
| "f1": 0.27 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2001, | |
| "recall": 0.1567, | |
| "fpr": 0.0674, | |
| "auroc": 0.6246, | |
| "pauc10": 0.1339, | |
| "precision": 0.9751, | |
| "f1": 0.27 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.2001, | |
| "recall": 0.1567, | |
| "fpr": 0.0674, | |
| "auroc": 0.6246, | |
| "pauc10": 0.1339, | |
| "precision": 0.9751, | |
| "f1": 0.27 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.6027, | |
| "recall": 0.2067, | |
| "fpr": 0.0013, | |
| "auroc": 0.9091, | |
| "pauc10": 0.6576, | |
| "precision": 0.9936, | |
| "f1": 0.3422 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.6027, | |
| "recall": 0.2067, | |
| "fpr": 0.0013, | |
| "auroc": 0.9091, | |
| "pauc10": 0.6576, | |
| "precision": 0.9936, | |
| "f1": 0.3422 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.6027, | |
| "recall": 0.2067, | |
| "fpr": 0.0013, | |
| "auroc": 0.9091, | |
| "pauc10": 0.6576, | |
| "precision": 0.9936, | |
| "f1": 0.3422 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.6027, | |
| "recall": 0.2067, | |
| "fpr": 0.0013, | |
| "auroc": 0.9091, | |
| "pauc10": 0.6576, | |
| "precision": 0.9936, | |
| "f1": 0.3422 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.913, | |
| "recall": 0.913 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.913, | |
| "recall": 0.913 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.913, | |
| "recall": 0.913 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.913, | |
| "recall": 0.913 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9266, | |
| "recall": 0.9036, | |
| "fpr": 0.041, | |
| "auroc": 0.9781, | |
| "pauc10": 0.8774, | |
| "precision": 0.9688, | |
| "f1": 0.9351 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.9266, | |
| "recall": 0.9036, | |
| "fpr": 0.041, | |
| "auroc": 0.9781, | |
| "pauc10": 0.8774, | |
| "precision": 0.9688, | |
| "f1": 0.9351 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9268, | |
| "recall": 0.904, | |
| "fpr": 0.041, | |
| "auroc": 0.9782, | |
| "pauc10": 0.8778, | |
| "precision": 0.9689, | |
| "f1": 0.9353 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.9268, | |
| "recall": 0.904, | |
| "fpr": 0.041, | |
| "auroc": 0.9782, | |
| "pauc10": 0.8778, | |
| "precision": 0.9689, | |
| "f1": 0.9353 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8953, | |
| "recall": 0.7926, | |
| "fpr": 0.0019, | |
| "auroc": 0.993, | |
| "pauc10": 0.9552, | |
| "precision": 0.9976, | |
| "f1": 0.8834 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8953, | |
| "recall": 0.7926, | |
| "fpr": 0.0019, | |
| "auroc": 0.993, | |
| "pauc10": 0.9552, | |
| "precision": 0.9976, | |
| "f1": 0.8834 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8953, | |
| "recall": 0.7926, | |
| "fpr": 0.0019, | |
| "auroc": 0.993, | |
| "pauc10": 0.9552, | |
| "precision": 0.9976, | |
| "f1": 0.8834 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.8953, | |
| "recall": 0.7926, | |
| "fpr": 0.0019, | |
| "auroc": 0.993, | |
| "pauc10": 0.9552, | |
| "precision": 0.9976, | |
| "f1": 0.8834 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7963, | |
| "recall": 0.7929, | |
| "fpr": 0.2, | |
| "auroc": 0.8544, | |
| "pauc10": 0.5338, | |
| "precision": 0.8073, | |
| "f1": 0.8 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7963, | |
| "recall": 0.7929, | |
| "fpr": 0.2, | |
| "auroc": 0.8544, | |
| "pauc10": 0.5338, | |
| "precision": 0.8073, | |
| "f1": 0.8 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7963, | |
| "recall": 0.7929, | |
| "fpr": 0.2, | |
| "auroc": 0.8544, | |
| "pauc10": 0.5338, | |
| "precision": 0.8073, | |
| "f1": 0.8 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7963, | |
| "recall": 0.7929, | |
| "fpr": 0.2, | |
| "auroc": 0.8544, | |
| "pauc10": 0.5338, | |
| "precision": 0.8073, | |
| "f1": 0.8 | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "neuraltrust-small": { | |
| "serve": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.7166, | |
| "fpr": 0.2834 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.7166, | |
| "fpr": 0.2834 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.72, | |
| "fpr": 0.28 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.72, | |
| "fpr": 0.28 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.452, | |
| "fpr": 0.548 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.452, | |
| "fpr": 0.548 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.4519, | |
| "fpr": 0.5481 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.4519, | |
| "fpr": 0.5481 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9191, | |
| "recall": 0.7895, | |
| "fpr": 0.0306, | |
| "auroc": 0.9737, | |
| "pauc10": 0.8142, | |
| "precision": 0.9091, | |
| "f1": 0.8451 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9191, | |
| "recall": 0.7895, | |
| "fpr": 0.0306, | |
| "auroc": 0.9737, | |
| "pauc10": 0.8142, | |
| "precision": 0.9091, | |
| "f1": 0.8451 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.8473, | |
| "recall": 0.7482, | |
| "fpr": 0.0407, | |
| "auroc": 0.9633, | |
| "pauc10": 0.7595, | |
| "precision": 0.9541, | |
| "f1": 0.8387 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.8473, | |
| "recall": 0.7482, | |
| "fpr": 0.0407, | |
| "auroc": 0.9633, | |
| "pauc10": 0.7595, | |
| "precision": 0.9541, | |
| "f1": 0.8387 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6034, | |
| "recall": 0.2333, | |
| "fpr": 0.0, | |
| "auroc": 0.842, | |
| "pauc10": 0.4613, | |
| "precision": 1.0, | |
| "f1": 0.3784 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6034, | |
| "recall": 0.2333, | |
| "fpr": 0.0, | |
| "auroc": 0.842, | |
| "pauc10": 0.4613, | |
| "precision": 1.0, | |
| "f1": 0.3784 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6034, | |
| "recall": 0.2333, | |
| "fpr": 0.0, | |
| "auroc": 0.842, | |
| "pauc10": 0.4613, | |
| "precision": 1.0, | |
| "f1": 0.3784 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.6034, | |
| "recall": 0.2333, | |
| "fpr": 0.0, | |
| "auroc": 0.842, | |
| "pauc10": 0.4613, | |
| "precision": 1.0, | |
| "f1": 0.3784 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7534, | |
| "recall": 0.7534 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.7534, | |
| "recall": 0.7534 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7552, | |
| "recall": 0.7552 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7552, | |
| "recall": 0.7552 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.8417, | |
| "recall": 0.8, | |
| "fpr": 0.1167, | |
| "auroc": 0.885, | |
| "pauc10": 0.5118, | |
| "precision": 0.8727, | |
| "f1": 0.8348 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.8417, | |
| "recall": 0.8, | |
| "fpr": 0.1167, | |
| "auroc": 0.885, | |
| "pauc10": 0.5118, | |
| "precision": 0.8727, | |
| "f1": 0.8348 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.8417, | |
| "recall": 0.8, | |
| "fpr": 0.1167, | |
| "auroc": 0.885, | |
| "pauc10": 0.5118, | |
| "precision": 0.8727, | |
| "f1": 0.8348 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.8417, | |
| "recall": 0.8, | |
| "fpr": 0.1167, | |
| "auroc": 0.885, | |
| "pauc10": 0.5118, | |
| "precision": 0.8727, | |
| "f1": 0.8348 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.0887, | |
| "recall": 0.0347, | |
| "fpr": 0.0, | |
| "auroc": 0.5412, | |
| "pauc10": 0.122, | |
| "precision": 1.0, | |
| "f1": 0.067 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.0887, | |
| "recall": 0.0347, | |
| "fpr": 0.0, | |
| "auroc": 0.5412, | |
| "pauc10": 0.122, | |
| "precision": 1.0, | |
| "f1": 0.067 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.0887, | |
| "recall": 0.0347, | |
| "fpr": 0.0, | |
| "auroc": 0.5412, | |
| "pauc10": 0.122, | |
| "precision": 1.0, | |
| "f1": 0.067 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.0887, | |
| "recall": 0.0347, | |
| "fpr": 0.0, | |
| "auroc": 0.5412, | |
| "pauc10": 0.122, | |
| "precision": 1.0, | |
| "f1": 0.067 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.536, | |
| "recall": 0.0733, | |
| "fpr": 0.0013, | |
| "auroc": 0.6098, | |
| "pauc10": 0.2449, | |
| "precision": 0.9821, | |
| "f1": 0.1365 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.536, | |
| "recall": 0.0733, | |
| "fpr": 0.0013, | |
| "auroc": 0.6098, | |
| "pauc10": 0.2449, | |
| "precision": 0.9821, | |
| "f1": 0.1365 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.536, | |
| "recall": 0.0733, | |
| "fpr": 0.0013, | |
| "auroc": 0.6098, | |
| "pauc10": 0.2449, | |
| "precision": 0.9821, | |
| "f1": 0.1365 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.536, | |
| "recall": 0.0733, | |
| "fpr": 0.0013, | |
| "auroc": 0.6098, | |
| "pauc10": 0.2449, | |
| "precision": 0.9821, | |
| "f1": 0.1365 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.218, | |
| "recall": 0.218 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.218, | |
| "recall": 0.218 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.218, | |
| "recall": 0.218 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.218, | |
| "recall": 0.218 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.8606, | |
| "recall": 0.7982, | |
| "fpr": 0.0513, | |
| "auroc": 0.934, | |
| "pauc10": 0.6738, | |
| "precision": 0.9564, | |
| "f1": 0.8702 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.8606, | |
| "recall": 0.7982, | |
| "fpr": 0.0513, | |
| "auroc": 0.934, | |
| "pauc10": 0.6738, | |
| "precision": 0.9564, | |
| "f1": 0.8702 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.8609, | |
| "recall": 0.7989, | |
| "fpr": 0.0513, | |
| "auroc": 0.9342, | |
| "pauc10": 0.6748, | |
| "precision": 0.9566, | |
| "f1": 0.8707 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.8609, | |
| "recall": 0.7989, | |
| "fpr": 0.0513, | |
| "auroc": 0.9342, | |
| "pauc10": 0.6748, | |
| "precision": 0.9566, | |
| "f1": 0.8707 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7344, | |
| "recall": 0.4833, | |
| "fpr": 0.0143, | |
| "auroc": 0.9387, | |
| "pauc10": 0.6695, | |
| "precision": 0.9713, | |
| "f1": 0.6455 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7344, | |
| "recall": 0.4833, | |
| "fpr": 0.0143, | |
| "auroc": 0.9387, | |
| "pauc10": 0.6695, | |
| "precision": 0.9713, | |
| "f1": 0.6455 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7344, | |
| "recall": 0.4833, | |
| "fpr": 0.0143, | |
| "auroc": 0.9387, | |
| "pauc10": 0.6695, | |
| "precision": 0.9713, | |
| "f1": 0.6455 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7344, | |
| "recall": 0.4833, | |
| "fpr": 0.0143, | |
| "auroc": 0.9387, | |
| "pauc10": 0.6695, | |
| "precision": 0.9713, | |
| "f1": 0.6455 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7486, | |
| "recall": 0.7071, | |
| "fpr": 0.2075, | |
| "auroc": 0.8224, | |
| "pauc10": 0.617, | |
| "precision": 0.7826, | |
| "f1": 0.743 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7486, | |
| "recall": 0.7071, | |
| "fpr": 0.2075, | |
| "auroc": 0.8224, | |
| "pauc10": 0.617, | |
| "precision": 0.7826, | |
| "f1": 0.743 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7486, | |
| "recall": 0.7071, | |
| "fpr": 0.2075, | |
| "auroc": 0.8224, | |
| "pauc10": 0.617, | |
| "precision": 0.7826, | |
| "f1": 0.743 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7486, | |
| "recall": 0.7071, | |
| "fpr": 0.2075, | |
| "auroc": 0.8224, | |
| "pauc10": 0.617, | |
| "precision": 0.7826, | |
| "f1": 0.743 | |
| } | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "notinject": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| }, | |
| "p05": { | |
| "n": 339, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.944, | |
| "fpr": 0.056 | |
| } | |
| } | |
| }, | |
| "xstest": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.7166, | |
| "fpr": 0.2834 | |
| }, | |
| "p05": { | |
| "n": 441, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.7166, | |
| "fpr": 0.2834 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.72, | |
| "fpr": 0.28 | |
| }, | |
| "p05": { | |
| "n": 450, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.72, | |
| "fpr": 0.28 | |
| } | |
| } | |
| }, | |
| "orbench_hard": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.452, | |
| "fpr": 0.548 | |
| }, | |
| "p05": { | |
| "n": 1312, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.452, | |
| "fpr": 0.548 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.4519, | |
| "fpr": 0.5481 | |
| }, | |
| "p05": { | |
| "n": 1319, | |
| "pos": 0, | |
| "thr": 0.5, | |
| "acc": 0.4519, | |
| "fpr": 0.5481 | |
| } | |
| } | |
| }, | |
| "jackhhao_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9265, | |
| "recall": 0.8158, | |
| "fpr": 0.0306, | |
| "auroc": 0.9742, | |
| "pauc10": 0.8142, | |
| "precision": 0.9118, | |
| "f1": 0.8611 | |
| }, | |
| "p05": { | |
| "n": 136, | |
| "pos": 38, | |
| "thr": 0.5, | |
| "acc": 0.9265, | |
| "recall": 0.8158, | |
| "fpr": 0.0306, | |
| "auroc": 0.9742, | |
| "pauc10": 0.8142, | |
| "precision": 0.9118, | |
| "f1": 0.8611 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.8511, | |
| "recall": 0.7554, | |
| "fpr": 0.0407, | |
| "auroc": 0.9613, | |
| "pauc10": 0.7375, | |
| "precision": 0.9545, | |
| "f1": 0.8434 | |
| }, | |
| "p05": { | |
| "n": 262, | |
| "pos": 139, | |
| "thr": 0.5, | |
| "acc": 0.8511, | |
| "recall": 0.7554, | |
| "fpr": 0.0407, | |
| "auroc": 0.9613, | |
| "pauc10": 0.7375, | |
| "precision": 0.9545, | |
| "f1": 0.8434 | |
| } | |
| } | |
| }, | |
| "deepset_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.5948, | |
| "recall": 0.2167, | |
| "fpr": 0.0, | |
| "auroc": 0.8417, | |
| "pauc10": 0.4613, | |
| "precision": 1.0, | |
| "f1": 0.3562 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.5948, | |
| "recall": 0.2167, | |
| "fpr": 0.0, | |
| "auroc": 0.8417, | |
| "pauc10": 0.4613, | |
| "precision": 1.0, | |
| "f1": 0.3562 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.5948, | |
| "recall": 0.2167, | |
| "fpr": 0.0, | |
| "auroc": 0.8417, | |
| "pauc10": 0.4613, | |
| "precision": 1.0, | |
| "f1": 0.3562 | |
| }, | |
| "p05": { | |
| "n": 116, | |
| "pos": 60, | |
| "thr": 0.5, | |
| "acc": 0.5948, | |
| "recall": 0.2167, | |
| "fpr": 0.0, | |
| "auroc": 0.8417, | |
| "pauc10": 0.4613, | |
| "precision": 1.0, | |
| "f1": 0.3562 | |
| } | |
| } | |
| }, | |
| "simsonsun_jailbreaks": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.746, | |
| "recall": 0.746 | |
| }, | |
| "p05": { | |
| "n": 2295, | |
| "pos": 2295, | |
| "thr": 0.5, | |
| "acc": 0.746, | |
| "recall": 0.746 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7469, | |
| "recall": 0.7469 | |
| }, | |
| "p05": { | |
| "n": 2426, | |
| "pos": 2426, | |
| "thr": 0.5, | |
| "acc": 0.7469, | |
| "recall": 0.7469 | |
| } | |
| } | |
| }, | |
| "boundary_pairs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.8417, | |
| "recall": 0.8, | |
| "fpr": 0.1167, | |
| "auroc": 0.8866, | |
| "pauc10": 0.5278, | |
| "precision": 0.8727, | |
| "f1": 0.8348 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.8417, | |
| "recall": 0.8, | |
| "fpr": 0.1167, | |
| "auroc": 0.8866, | |
| "pauc10": 0.5278, | |
| "precision": 0.8727, | |
| "f1": 0.8348 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.8417, | |
| "recall": 0.8, | |
| "fpr": 0.1167, | |
| "auroc": 0.8866, | |
| "pauc10": 0.5278, | |
| "precision": 0.8727, | |
| "f1": 0.8348 | |
| }, | |
| "p05": { | |
| "n": 240, | |
| "pos": 120, | |
| "thr": 0.5, | |
| "acc": 0.8417, | |
| "recall": 0.8, | |
| "fpr": 0.1167, | |
| "auroc": 0.8866, | |
| "pauc10": 0.5278, | |
| "precision": 0.8727, | |
| "f1": 0.8348 | |
| } | |
| } | |
| }, | |
| "bipia": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.09, | |
| "recall": 0.036, | |
| "fpr": 0.0, | |
| "auroc": 0.5246, | |
| "pauc10": 0.1152, | |
| "precision": 1.0, | |
| "f1": 0.0695 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.09, | |
| "recall": 0.036, | |
| "fpr": 0.0, | |
| "auroc": 0.5246, | |
| "pauc10": 0.1152, | |
| "precision": 1.0, | |
| "f1": 0.0695 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.09, | |
| "recall": 0.036, | |
| "fpr": 0.0, | |
| "auroc": 0.5246, | |
| "pauc10": 0.1152, | |
| "precision": 1.0, | |
| "f1": 0.0695 | |
| }, | |
| "p05": { | |
| "n": 3178, | |
| "pos": 3000, | |
| "thr": 0.5, | |
| "acc": 0.09, | |
| "recall": 0.036, | |
| "fpr": 0.0, | |
| "auroc": 0.5246, | |
| "pauc10": 0.1152, | |
| "precision": 1.0, | |
| "f1": 0.0695 | |
| } | |
| } | |
| }, | |
| "piarena": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5333, | |
| "recall": 0.0667, | |
| "fpr": 0.0, | |
| "auroc": 0.6292, | |
| "pauc10": 0.2649, | |
| "precision": 1.0, | |
| "f1": 0.125 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5333, | |
| "recall": 0.0667, | |
| "fpr": 0.0, | |
| "auroc": 0.6292, | |
| "pauc10": 0.2649, | |
| "precision": 1.0, | |
| "f1": 0.125 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5333, | |
| "recall": 0.0667, | |
| "fpr": 0.0, | |
| "auroc": 0.6292, | |
| "pauc10": 0.2649, | |
| "precision": 1.0, | |
| "f1": 0.125 | |
| }, | |
| "p05": { | |
| "n": 1500, | |
| "pos": 750, | |
| "thr": 0.5, | |
| "acc": 0.5333, | |
| "recall": 0.0667, | |
| "fpr": 0.0, | |
| "auroc": 0.6292, | |
| "pauc10": 0.2649, | |
| "precision": 1.0, | |
| "f1": 0.125 | |
| } | |
| } | |
| }, | |
| "llmail_phase2": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.1815, | |
| "recall": 0.1815 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.1815, | |
| "recall": 0.1815 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.1815, | |
| "recall": 0.1815 | |
| }, | |
| "p05": { | |
| "n": 2000, | |
| "pos": 2000, | |
| "thr": 0.5, | |
| "acc": 0.1815, | |
| "recall": 0.1815 | |
| } | |
| } | |
| }, | |
| "neuralchemy_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.8564, | |
| "recall": 0.7909, | |
| "fpr": 0.0513, | |
| "auroc": 0.9331, | |
| "pauc10": 0.6691, | |
| "precision": 0.956, | |
| "f1": 0.8657 | |
| }, | |
| "p05": { | |
| "n": 940, | |
| "pos": 550, | |
| "thr": 0.5, | |
| "acc": 0.8564, | |
| "recall": 0.7909, | |
| "fpr": 0.0513, | |
| "auroc": 0.9331, | |
| "pauc10": 0.6691, | |
| "precision": 0.956, | |
| "f1": 0.8657 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "recall": 0.7917, | |
| "fpr": 0.0513, | |
| "auroc": 0.9333, | |
| "pauc10": 0.67, | |
| "precision": 0.9562, | |
| "f1": 0.8662 | |
| }, | |
| "p05": { | |
| "n": 942, | |
| "pos": 552, | |
| "thr": 0.5, | |
| "acc": 0.8567, | |
| "recall": 0.7917, | |
| "fpr": 0.0513, | |
| "auroc": 0.9333, | |
| "pauc10": 0.67, | |
| "precision": 0.9562, | |
| "f1": 0.8662 | |
| } | |
| } | |
| }, | |
| "slabs_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7344, | |
| "recall": 0.4833, | |
| "fpr": 0.0143, | |
| "auroc": 0.9387, | |
| "pauc10": 0.6695, | |
| "precision": 0.9713, | |
| "f1": 0.6455 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7344, | |
| "recall": 0.4833, | |
| "fpr": 0.0143, | |
| "auroc": 0.9387, | |
| "pauc10": 0.6695, | |
| "precision": 0.9713, | |
| "f1": 0.6455 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7344, | |
| "recall": 0.4833, | |
| "fpr": 0.0143, | |
| "auroc": 0.9387, | |
| "pauc10": 0.6695, | |
| "precision": 0.9713, | |
| "f1": 0.6455 | |
| }, | |
| "p05": { | |
| "n": 2101, | |
| "pos": 1051, | |
| "thr": 0.5, | |
| "acc": 0.7344, | |
| "recall": 0.4833, | |
| "fpr": 0.0143, | |
| "auroc": 0.9387, | |
| "pauc10": 0.6695, | |
| "precision": 0.9713, | |
| "f1": 0.6455 | |
| } | |
| } | |
| }, | |
| "agentic5k_test": { | |
| "decontaminated": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7486, | |
| "recall": 0.7071, | |
| "fpr": 0.2075, | |
| "auroc": 0.8224, | |
| "pauc10": 0.617, | |
| "precision": 0.7826, | |
| "f1": 0.743 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7486, | |
| "recall": 0.7071, | |
| "fpr": 0.2075, | |
| "auroc": 0.8224, | |
| "pauc10": 0.617, | |
| "precision": 0.7826, | |
| "f1": 0.743 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7486, | |
| "recall": 0.7071, | |
| "fpr": 0.2075, | |
| "auroc": 0.8224, | |
| "pauc10": 0.617, | |
| "precision": 0.7826, | |
| "f1": 0.743 | |
| }, | |
| "p05": { | |
| "n": 545, | |
| "pos": 280, | |
| "thr": 0.5, | |
| "acc": 0.7486, | |
| "recall": 0.7071, | |
| "fpr": 0.2075, | |
| "auroc": 0.8224, | |
| "pauc10": 0.617, | |
| "precision": 0.7826, | |
| "f1": 0.743 | |
| } | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "macro_card10": { | |
| "galvanize-2": { | |
| "serve": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.8695, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8936, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.8717, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8952, | |
| "n_sets": 10 | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.8688, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8918, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.8683, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8924, | |
| "n_sets": 10 | |
| } | |
| } | |
| } | |
| }, | |
| "horizon-small": { | |
| "serve": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.8866, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8866, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.8893, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8893, | |
| "n_sets": 10 | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.8864, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8864, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.8871, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8871, | |
| "n_sets": 10 | |
| } | |
| } | |
| } | |
| }, | |
| "horizon-base": { | |
| "serve": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.9017, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.9017, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.9039, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.9039, | |
| "n_sets": 10 | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.9029, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.9029, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.9031, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.9031, | |
| "n_sets": 10 | |
| } | |
| } | |
| } | |
| }, | |
| "horizon-large": { | |
| "serve": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.8839, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8839, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.8846, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8846, | |
| "n_sets": 10 | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.8992, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8992, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.899, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.899, | |
| "n_sets": 10 | |
| } | |
| } | |
| } | |
| }, | |
| "galvanize-60m": { | |
| "serve": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.5213, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.5304, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.5307, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.5421, | |
| "n_sets": 10 | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.5954, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.6018, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.6057, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.6136, | |
| "n_sets": 10 | |
| } | |
| } | |
| } | |
| }, | |
| "protectai-v2": { | |
| "serve": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.6218, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.6218, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.6257, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.6257, | |
| "n_sets": 10 | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.6306, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.6306, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.6349, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.6349, | |
| "n_sets": 10 | |
| } | |
| } | |
| } | |
| }, | |
| "deepset": { | |
| "serve": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.7838, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.7838, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.8092, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8092, | |
| "n_sets": 10 | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.7838, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.7838, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.8092, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.8092, | |
| "n_sets": 10 | |
| } | |
| } | |
| } | |
| }, | |
| "wolf-defender": { | |
| "serve": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.7652, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.7652, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.7708, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.7708, | |
| "n_sets": 10 | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.7542, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.7542, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.7589, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.7589, | |
| "n_sets": 10 | |
| } | |
| } | |
| } | |
| }, | |
| "neuraltrust-small": { | |
| "serve": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.5346, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.5346, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.5344, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.5344, | |
| "n_sets": 10 | |
| } | |
| } | |
| }, | |
| "hz": { | |
| "decontaminated": { | |
| "own": { | |
| "value": 0.5287, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.5287, | |
| "n_sets": 10 | |
| } | |
| }, | |
| "full": { | |
| "own": { | |
| "value": 0.5273, | |
| "n_sets": 10 | |
| }, | |
| "p05": { | |
| "value": 0.5273, | |
| "n_sets": 10 | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "reproduction_vs_horizon_published": { | |
| "deepset": { | |
| "delta_acc": { | |
| "agentic5k_test": 0.0, | |
| "bipia": 0.0, | |
| "boundary_pairs_test": 0.0, | |
| "deepset_test": 0.0, | |
| "jackhhao_test": 0.0, | |
| "llmail_phase2": 0.0, | |
| "neuralchemy_test": 0.0, | |
| "notinject": 0.0, | |
| "orbench_hard": 0.0, | |
| "piarena": 0.0, | |
| "simsonsun_jailbreaks": -0.0004, | |
| "slabs_test": 0.0019, | |
| "xstest": 0.0 | |
| }, | |
| "delta_auroc": { | |
| "agentic5k_test": -0.0006, | |
| "bipia": 0.0008, | |
| "boundary_pairs_test": 0.002, | |
| "deepset_test": 0.0, | |
| "jackhhao_test": 0.0019, | |
| "neuralchemy_test": -0.0001, | |
| "piarena": 0.0001, | |
| "slabs_test": 0.0 | |
| }, | |
| "max_abs_delta_acc": 0.0019, | |
| "max_abs_delta_auroc": 0.002 | |
| }, | |
| "horizon-base": { | |
| "delta_acc": { | |
| "agentic5k_test": -0.0019, | |
| "bipia": 0.0022, | |
| "boundary_pairs_test": 0.0, | |
| "deepset_test": 0.0086, | |
| "jackhhao_test": 0.0, | |
| "llmail_phase2": 0.0, | |
| "neuralchemy_test": 0.0, | |
| "notinject": 0.0, | |
| "orbench_hard": -0.0007, | |
| "piarena": 0.0007, | |
| "simsonsun_jailbreaks": 0.0013, | |
| "slabs_test": 0.0005, | |
| "xstest": 0.0 | |
| }, | |
| "delta_auroc": { | |
| "agentic5k_test": 0.0, | |
| "bipia": 0.0001, | |
| "boundary_pairs_test": -0.0001, | |
| "deepset_test": 0.0005, | |
| "jackhhao_test": 0.0, | |
| "neuralchemy_test": 0.0, | |
| "piarena": 0.0, | |
| "slabs_test": 0.0 | |
| }, | |
| "max_abs_delta_acc": 0.0086, | |
| "max_abs_delta_auroc": 0.0005 | |
| }, | |
| "neuraltrust-small": { | |
| "delta_acc": { | |
| "agentic5k_test": -0.0055, | |
| "bipia": 0.0003, | |
| "boundary_pairs_test": 0.0, | |
| "deepset_test": 0.0086, | |
| "jackhhao_test": 0.0, | |
| "llmail_phase2": -0.001, | |
| "neuralchemy_test": -0.0021, | |
| "notinject": 0.0, | |
| "orbench_hard": -0.0007, | |
| "piarena": 0.0, | |
| "simsonsun_jailbreaks": 0.0004, | |
| "slabs_test": -0.0019, | |
| "xstest": 0.0 | |
| }, | |
| "delta_auroc": { | |
| "agentic5k_test": 0.0018, | |
| "bipia": 0.0004, | |
| "boundary_pairs_test": -0.0007, | |
| "deepset_test": -0.0015, | |
| "jackhhao_test": -0.0001, | |
| "neuralchemy_test": 0.0, | |
| "piarena": 0.0004, | |
| "slabs_test": 0.0 | |
| }, | |
| "max_abs_delta_acc": 0.0086, | |
| "max_abs_delta_auroc": 0.0018 | |
| }, | |
| "protectai-v2": { | |
| "delta_acc": { | |
| "agentic5k_test": 0.0, | |
| "bipia": 0.0003, | |
| "boundary_pairs_test": 0.0, | |
| "deepset_test": 0.0, | |
| "jackhhao_test": 0.0, | |
| "llmail_phase2": -0.0005, | |
| "neuralchemy_test": 0.0, | |
| "notinject": 0.003, | |
| "orbench_hard": 0.0, | |
| "piarena": 0.0, | |
| "simsonsun_jailbreaks": 0.0008, | |
| "slabs_test": 0.0005, | |
| "xstest": 0.0 | |
| }, | |
| "delta_auroc": { | |
| "agentic5k_test": -0.0003, | |
| "bipia": -0.0002, | |
| "boundary_pairs_test": 0.0007, | |
| "deepset_test": 0.0009, | |
| "jackhhao_test": -0.0002, | |
| "neuralchemy_test": 0.0004, | |
| "piarena": 0.0001, | |
| "slabs_test": 0.0 | |
| }, | |
| "max_abs_delta_acc": 0.003, | |
| "max_abs_delta_auroc": 0.0009 | |
| }, | |
| "wolf-defender": { | |
| "delta_acc": { | |
| "agentic5k_test": 0.0055, | |
| "bipia": -0.001, | |
| "boundary_pairs_test": -0.0042, | |
| "deepset_test": 0.0, | |
| "jackhhao_test": 0.0, | |
| "llmail_phase2": 0.0005, | |
| "neuralchemy_test": -0.001, | |
| "notinject": 0.0029, | |
| "orbench_hard": 0.0023, | |
| "piarena": 0.0014, | |
| "simsonsun_jailbreaks": -0.0016, | |
| "slabs_test": 0.0005, | |
| "xstest": 0.0 | |
| }, | |
| "delta_auroc": { | |
| "agentic5k_test": 0.0001, | |
| "bipia": -0.0003, | |
| "boundary_pairs_test": 0.0005, | |
| "deepset_test": -0.0013, | |
| "jackhhao_test": -0.0003, | |
| "neuralchemy_test": -0.0001, | |
| "piarena": 0.0002, | |
| "slabs_test": -0.0001 | |
| }, | |
| "max_abs_delta_acc": 0.0055, | |
| "max_abs_delta_auroc": 0.0013 | |
| } | |
| }, | |
| "not_evaluated": { | |
| "leolee99/PIGuard": "custom architecture that needs trust_remote_code (excluded by policy)", | |
| "meta-llama/Llama-Prompt-Guard-2-86M": "gated (manual approval, Llama license)", | |
| "meta-llama/Llama-Prompt-Guard-2-22M": "gated (manual approval, Llama license)", | |
| "qualifire/prompt-injections-benchmark (dataset)": "gated and non-commercial; not in the public suite", | |
| "Mindgard/evaded-prompt-injection-and-jailbreak-samples (dataset)": "gated and non-commercial; not in the public suite", | |
| "Horizon synthetic_* sets (dataset)": "generated with the same generator as Horizon's training data (in-distribution for Horizon and, through initialisation, for Galvanize 2)", | |
| "Commercial APIs (e.g. Lakera Guard, Azure Prompt Shields)": "paid APIs, not open weights" | |
| } | |
| } | |