Add concise experiment navigation and task-level reproducibility manifests

latency-aware/deadly-corridor/small-policy/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0/artifact.json

@@ -7,29 +7,34 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/deadly-corridor/small-policy/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0/train/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0/checkpoint_p0/checkpoint_000048836_25004032.pth"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/deadly-corridor/small-policy/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0/train/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0/train/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0/git.diff"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0/train/memory_rollout_teacher:deadly_corridor:teacher:fixed_l6:fs4:obs8p75:stride4:seed0/sf_log.txt"
     }
   ],
   "asset_state": "teacher_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "run": "memory/rollout_teachers/deadly_corridor",

latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_kv_memory_flex/artifact.json

@@ -7,39 +7,44 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_kv_memory_flex/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_kv_memory_flex"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_kv_memory_flex/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_kv_memory_flex/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_kv_memory_flex/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_kv_memory_flex/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_kv_memory_flex/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "deadly_corridor_fix_latency_6_1000ep_7k2steps_kv_memory_flex",
@@ -67,7 +72,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/deadly-corridor/deadly_corridor_fixed_latency_6_1000ep_7k2steps",
       "config_name": "deadly_corridor_fixed_latency_6_1000ep_7k2steps"
     }

latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_plain_multi/artifact.json

@@ -7,39 +7,44 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_plain_multi/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_plain_multi"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_plain_multi/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_plain_multi/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_plain_multi/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_plain_multi/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_plain_multi/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "deadly_corridor_fix_latency_6_1000ep_7k2steps_plain_multi",
@@ -67,7 +72,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/deadly-corridor/deadly_corridor_fixed_latency_6_1000ep_7k2steps",
       "config_name": "deadly_corridor_fixed_latency_6_1000ep_7k2steps"
     }

latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline/artifact.json

@@ -7,49 +7,54 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline/eval/post_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline/eval_inputs/deadly_corridor_fixed_latency_6.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "deadly_corridor_fix_latency_6_1000ep_7k2steps_single_baseline",
@@ -77,7 +82,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/deadly-corridor/deadly_corridor_fixed_latency_6_1000ep_7k2steps",
       "config_name": "deadly_corridor_fixed_latency_6_1000ep_7k2steps"
     }

latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_stitch/artifact.json

@@ -7,39 +7,44 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_stitch/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/deadly_corridor_fix_latency_6_1000ep_7k2steps_stitch"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_stitch/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_stitch/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_stitch/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_stitch/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/deadly_corridor_fix_latency_6_1000ep_7k2steps_stitch/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "deadly_corridor_fix_latency_6_1000ep_7k2steps_stitch",
@@ -67,7 +72,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/deadly-corridor/deadly_corridor_fixed_latency_6_1000ep_7k2steps",
       "config_name": "deadly_corridor_fixed_latency_6_1000ep_7k2steps"
     }

latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/openvla_deadly_corridor_fix_latency_2/artifact.json

@@ -7,18 +7,23 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/openvla_deadly_corridor_fix_latency_2/checkpoints/best_model.safetensors"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/openvla_deadly_corridor_fix_latency_2/checkpoints/steps_500_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/deadly-corridor/vla/starvla-qwenoft-h1/openvla_deadly_corridor_fix_latency_2"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_deadly_corridor_fix_latency_2",
@@ -34,7 +39,7 @@
       "training_data_mix": "deadly_corridor_train__bridge",
       "dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/deadly-corridor/deadly_corridor_fix_latency_2",
         "evidence": "Training config uses the exact source dataset name as its local data-root directory. Original metadata and statistics are retained separately."
       },
@@ -78,7 +83,7 @@
       "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
       "training_recipe_dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/deadly-corridor/deadly_corridor_fix_latency_2",
         "evidence": "Training config uses the exact source dataset name as its local data-root directory. Original metadata and statistics are retained separately."
       }

latency-aware/deadly-corridor/vla/starvla-wanoft-h8/wan_oft_deadly_corridor_fix_latency_6_context5_standard_sft_1000_effbs128_224_currentbce/artifact.json

@@ -7,29 +7,34 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/deadly-corridor/vla/starvla-wanoft-h8/wan_oft_deadly_corridor_fix_latency_6_context5_standard_sft_1000_effbs128_224_currentbce/checkpoints/steps_1000_pytorch_model.pt"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/deadly-corridor/vla/starvla-wanoft-h8/wan_oft_deadly_corridor_fix_latency_6_context5_standard_sft_1000_effbs128_224_currentbce"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/wan_oft_deadly_corridor_fix_latency_6_context5_standard_sft_1000_effbs128_224_currentbce/eval/fixed_6/eval_latency_6/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/wan_oft_deadly_corridor_fix_latency_6_context5_standard_sft_1000_effbs128_224_currentbce/eval/fixed_6/queue_eval_latency_6.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/deadly-corridor/wan_oft_deadly_corridor_fix_latency_6_context5_standard_sft_1000_effbs128_224_currentbce/eval/fixed_6/queue_eval_results.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "wan_oft_deadly_corridor_fix_latency_6_context5_standard_sft_1000_effbs128_224_currentbce",
@@ -57,7 +62,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/deadly-corridor/deadly_corridor_fixed_latency_6_1000ep_7k2steps",
       "config_name": "deadly_corridor_fixed_latency_6_1000ep_7k2steps"
     }

latency-aware/demon-attack/small-policy/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0/artifact.json

@@ -7,29 +7,34 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/small-policy/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0/train/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0/checkpoint_p0/checkpoint_000048880_25026560.pth"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/small-policy/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0/train/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0/train/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0/git.diff"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0/train/memory_rollout_teacher:demon_attack:teacher:fixed_l6:fs4:obs15:stride4:seed0/sf_log.txt"
     }
   ],
   "asset_state": "teacher_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "run": "memory/rollout_teachers/demon_attack",

latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_2_200ep/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_2_200ep/checkpoints/steps_7000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_2_200ep"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_demon_attack_200ep",
@@ -65,9 +70,23 @@
         "paper/figures/fig6_latency_robustness/demon_attack"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/demon-attack/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/demon-attack/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "demon_attack_fix_latency_2_200ep",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_4_200ep/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_4_200ep/checkpoints/steps_7000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_4_200ep"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_demon_attack_200ep",
@@ -65,9 +70,23 @@
         "paper/figures/fig6_latency_robustness/demon_attack"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/demon-attack/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/demon-attack/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "demon_attack_fix_latency_4_200ep",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep/checkpoints/steps_7000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_demon_attack_200ep",
@@ -65,9 +70,23 @@
         "paper/figures/fig6_latency_robustness/demon_attack"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/demon-attack/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/demon-attack/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "demon_attack_fix_latency_6_200ep",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex/artifact.json

@@ -7,44 +7,49 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex/eval/post_train/step_4000.cap3600.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "demon_attack_fix_latency_6_200ep_7k2steps_kv_memory_flex",
@@ -72,7 +77,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps",
       "config_name": "demon_attack_fixed_latency_6_200ep_7k2steps"
     }

latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi/artifact.json

@@ -7,49 +7,54 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi/eval/post_train/step_4000.cap3600.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi/eval/post_train/step_4000.pre_framealign.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_plain_multi/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "demon_attack_fix_latency_6_200ep_7k2steps_plain_multi",
@@ -77,7 +82,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps",
       "config_name": "demon_attack_fixed_latency_6_200ep_7k2steps"
     }

latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_single_baseline/artifact.json

@@ -7,44 +7,49 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_single_baseline/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_single_baseline"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_single_baseline/eval/post_train/step_4000.cap3600.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_single_baseline/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_single_baseline/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_single_baseline/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_single_baseline/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_single_baseline/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "demon_attack_fix_latency_6_200ep_7k2steps_single_baseline",
@@ -72,7 +77,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps",
       "config_name": "demon_attack_fixed_latency_6_200ep_7k2steps"
     }

latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_stitch/artifact.json

@@ -7,49 +7,54 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_stitch/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_6_200ep_7k2steps_stitch"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_stitch/eval/post_train/step_4000.cap3600.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_stitch/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_stitch/eval/post_train/step_4000.pre_framealign.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_stitch/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_stitch/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_stitch/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/demon_attack_fix_latency_6_200ep_7k2steps_stitch/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "demon_attack_fix_latency_6_200ep_7k2steps_stitch",
@@ -77,7 +82,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps",
       "config_name": "demon_attack_fixed_latency_6_200ep_7k2steps"
     }

latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_8_200ep/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_8_200ep/checkpoints/steps_7000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/demon_attack_fix_latency_8_200ep"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_demon_attack_200ep",
@@ -65,9 +70,23 @@
         "paper/figures/fig6_latency_robustness/demon_attack"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/demon-attack/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/demon-attack/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "demon_attack_fix_latency_8_200ep",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/artifact.json

@@ -7,44 +7,49 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/eval/mid_train/step_5000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/eval/post_train/step_5000.json"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15",
@@ -60,7 +65,7 @@
       "training_data_mix": "demon_attack_train__bridge",
       "dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15",
         "source_subdir": null
       },
@@ -106,7 +111,7 @@
       "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
       "training_recipe_dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15",
         "source_subdir": null
       }
@@ -115,7 +120,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15"
     }
   ]

latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2/artifact.json

@@ -5,9 +5,14 @@
     "path": "openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2"
   },
   "checkpoint": [],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2"
+  },
   "evaluation": [],
   "asset_state": "metadata_only",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2",
@@ -71,7 +76,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2"
     }
   ]

latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/artifact.json

@@ -7,54 +7,59 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/eval/post_train/step_4000.max3600.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/eval/post_train/step_4000.part0of2.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3/eval/post_train/step_4000.part1of2.json"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2_exp3",
@@ -125,7 +130,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_2"
     }
   ]

latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2/README.md

@@ -4,4 +4,4 @@ Demon Attack · qwenoft · H1
 
 [Experiment settings and asset manifest](https://huggingface.co/datasets/latency-sensitive-bench/benchmark-datasets/tree/main/experiments/visual-history/archive/demon-attack)
 
-Download this run at the fixed revision in the task manifest. Keep the checkpoint, original configuration and normalization statistics together. Original local paths describe source provenance. `artifact.json` records the current asset references.
+This historical run shares identical inference weights with another retained run. Download this run’s own config and statistics, and place the canonical checkpoint referenced in `artifact.json` under its local `checkpoints/` directory before native loading. Distinct source run and seed metadata remain preserved. Original local paths describe source provenance. `artifact.json` records the current asset references.

latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2/artifact.json

@@ -7,44 +7,49 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
-      "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2/checkpoints/steps_5000_model.safetensors"
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+      "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2/eval/mid_train/step_5000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2/eval/post_train/step_5000.json"
     }
   ],
-  "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "asset_state": "shared_checkpoint",
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_exp2",
@@ -60,7 +65,7 @@
       "training_data_mix": "demon_attack_train__bridge",
       "dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15",
         "source_subdir": null
       },
@@ -106,7 +111,7 @@
       "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
       "training_recipe_dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15",
         "source_subdir": null
       }
@@ -115,7 +120,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15"
     }
   ]

latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1/README.md

@@ -4,4 +4,4 @@ Demon Attack · qwenoft · H1
 
 [Experiment settings and asset manifest](https://huggingface.co/datasets/latency-sensitive-bench/benchmark-datasets/tree/main/experiments/visual-history/archive/demon-attack)
 
-Download this run at the fixed revision in the task manifest. Keep the checkpoint, original configuration and normalization statistics together. Original local paths describe source provenance. `artifact.json` records the current asset references.
+This historical run shares identical inference weights with another retained run. Download this run’s own config and statistics, and place the canonical checkpoint referenced in `artifact.json` under its local `checkpoints/` directory before native loading. Distinct source run and seed metadata remain preserved. Original local paths describe source provenance. `artifact.json` records the current asset references.

latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1/artifact.json

@@ -7,44 +7,49 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
-      "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1/checkpoints/steps_5000_model.safetensors"
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+      "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1/eval/mid_train/step_5000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1/eval/post_train/step_5000.json"
     }
   ],
-  "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "asset_state": "shared_checkpoint",
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_bridge_demon_attack_fixed_latency_6_200ep_7k2steps_ghost15_seed1",
@@ -60,7 +65,7 @@
       "training_data_mix": "demon_attack_train__bridge",
       "dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15",
         "source_subdir": null
       },
@@ -106,7 +111,7 @@
       "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
       "training_recipe_dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15",
         "source_subdir": null
       }
@@ -115,7 +120,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps_ghost15"
     }
   ]

latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_demon_attack_fix_latency_2_small/artifact.json

@@ -7,18 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
-      "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_demon_attack_fix_latency_2_small/checkpoints/best_model.safetensors"
-    },
-    {
-      "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_demon_attack_fix_latency_2_small/checkpoints/steps_2000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-qwenoft-h1/openvla_demon_attack_fix_latency_2_small"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_demon_attack_fix_latency_2_small",

latency-aware/demon-attack/vla/starvla-wanoft-h8/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce/artifact.json

@@ -7,49 +7,54 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/demon-attack/vla/starvla-wanoft-h8/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce/checkpoints/steps_5000_pytorch_model.pt"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/demon-attack/vla/starvla-wanoft-h8/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce/eval/fixed_6/eval_latency_6/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce/eval/fixed_6/queue_eval_latency_6.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce/eval/fixed_6/queue_eval_results.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/demon-attack/wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "wan_oft_demon_attack_fix_latency_6_context5_standard_sft_5000_effbs128_224_currentce",
@@ -77,7 +82,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/demon-attack/demon_attack_fixed_latency_6_200ep_7k2steps",
       "config_name": "demon_attack_fixed_latency_6_200ep_7k2steps"
     }

latency-aware/flappy/small-policy/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0/artifact.json

@@ -7,29 +7,34 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/small-policy/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0/train/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0/checkpoint_p0/checkpoint_000048840_25034752.pth"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/small-policy/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0/train/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0/train/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0/git.diff"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0/train/memory_rollout_teacher:flappy:teacher:fixed_l3:fs4:obs30:stride1:seed0/sf_log.txt"
     }
   ],
   "asset_state": "teacher_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "run": "memory/rollout_teachers/flappy",

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_curriculum_cumulative_40ep_per_latency/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_curriculum_cumulative_40ep_per_latency/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_curriculum_cumulative_40ep_per_latency"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_flappy_200ep",
@@ -65,9 +70,23 @@
         "paper/figures/fig6_latency_robustness/flappy"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "flappy_curriculum_cumulative_40ep_per_latency",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_curriculum_exclusive_40ep_per_latency/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_curriculum_exclusive_40ep_per_latency/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_curriculum_exclusive_40ep_per_latency"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_flappy_200ep",
@@ -65,9 +70,23 @@
         "paper/figures/fig6_latency_robustness/flappy"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "flappy_curriculum_exclusive_40ep_per_latency",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_1_200ep/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_1_200ep/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_1_200ep"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_flappy_200ep",
@@ -63,9 +68,23 @@
         "paper/figures/fig6_latency_robustness/flappy"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "flappy_fix_latency_1_200ep",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_flappy_200ep",
@@ -64,9 +69,23 @@
         "paper/figures/fig6_latency_robustness/flappy"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "flappy_fix_latency_2_200ep",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_kv_memory/artifact.json

@@ -7,204 +7,209 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_kv_memory/checkpoints/steps_4000_pytorch_model.pt"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_kv_memory"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_1250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_1750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_2250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_2500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_2750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_3250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_3500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_3750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/mid_train/step_750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_1000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_1250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_1500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_1750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_2000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_2250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_2500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_2750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_3000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_3250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_3500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_3750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/mid_train/step_750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/latency_bench_eval/post_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "run": "flappy_fix_latency_2_200ep_7k2steps_kv_memory",
@@ -232,7 +237,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fix_latency_2_200ep_7k2steps",
       "config_name": "flappy_fix_latency_2_200ep_7k2steps"
     }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/artifact.json

@@ -7,374 +7,379 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_1250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_1750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_2250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_2500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_2750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_3250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_3500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_3750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/post_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/post_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/latency_bench_eval/post_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "run": "flappy_fix_latency_2_200ep_7k2steps_kv_memory_flex",
@@ -402,7 +407,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fix_latency_2_200ep_7k2steps",
       "config_name": "flappy_fix_latency_2_200ep_7k2steps"
     }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_plain_multi/artifact.json

@@ -7,374 +7,379 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_plain_multi/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_plain_multi"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_1250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_1750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_2250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_2500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_2750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_3250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_3500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_3750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/mid_train/step_750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/post_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/post_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/latency_bench_eval/post_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_plain_multi/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "run": "flappy_fix_latency_2_200ep_7k2steps_plain_multi",
@@ -402,7 +407,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fix_latency_2_200ep_7k2steps",
       "config_name": "flappy_fix_latency_2_200ep_7k2steps"
     }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_single_baseline/artifact.json

@@ -7,374 +7,379 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_single_baseline/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_single_baseline"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_1250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_1750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_2250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_2500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_2750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_3250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_3500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_3750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/mid_train/step_750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/post_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/post_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/latency_bench_eval/post_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_single_baseline/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "run": "flappy_fix_latency_2_200ep_7k2steps_single_baseline",
@@ -402,7 +407,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fix_latency_2_200ep_7k2steps",
       "config_name": "flappy_fix_latency_2_200ep_7k2steps"
     }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_stitch/artifact.json

@@ -7,374 +7,379 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_stitch/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_2_200ep_7k2steps_stitch"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_1250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_1750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_2250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_2500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_2750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_3250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_3500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_3750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/mid_train/step_750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3250/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_500/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_750/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/post_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/post_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/latency_bench_eval/post_train/step_4000/flappy/latency_2/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_2_200ep_7k2steps_stitch/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "run": "flappy_fix_latency_2_200ep_7k2steps_stitch",
@@ -402,7 +407,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fix_latency_2_200ep_7k2steps",
       "config_name": "flappy_fix_latency_2_200ep_7k2steps"
     }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_flappy_200ep",
@@ -64,9 +69,23 @@
         "paper/figures/fig6_latency_robustness/flappy"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "flappy_fix_latency_3_200ep",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/artifact.json

@@ -7,374 +7,379 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_1250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_1750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_2250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_2500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_2750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_3250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_3500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_3750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/mid_train/step_750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_1750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_2750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_3750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_4000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/mid_train/step_750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/post_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/post_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/latency_bench_eval/post_train/step_4000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "flappy_fix_latency_3_200ep_7k2steps_kv_memory_flex",
@@ -402,7 +407,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps",
       "config_name": "flappy_fixed_latency_3_200ep_7k2steps"
     }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_plain_multi/artifact.json

@@ -7,374 +7,379 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_plain_multi/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_plain_multi"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_1250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_1750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_2250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_2500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_2750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_3250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_3500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_3750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/mid_train/step_750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_1750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_2750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_3750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_4000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/mid_train/step_750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/post_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/post_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/latency_bench_eval/post_train/step_4000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_plain_multi/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "flappy_fix_latency_3_200ep_7k2steps_plain_multi",
@@ -402,7 +407,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps",
       "config_name": "flappy_fixed_latency_3_200ep_7k2steps"
     }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_single_baseline/artifact.json

@@ -7,374 +7,379 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_single_baseline/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_single_baseline"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_1250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_1750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_2250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_2500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_2750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_3250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_3500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_3750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/mid_train/step_750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_1750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_2750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_3750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_4000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/mid_train/step_750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/post_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/post_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/latency_bench_eval/post_train/step_4000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_single_baseline/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "flappy_fix_latency_3_200ep_7k2steps_single_baseline",
@@ -402,7 +407,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps",
       "config_name": "flappy_fixed_latency_3_200ep_7k2steps"
     }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_stitch/artifact.json

@@ -7,384 +7,389 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_stitch/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_3_200ep_7k2steps_stitch"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_1000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_1250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_1750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_2000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_2250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_2500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_2750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_3000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_3250.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_3500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_3750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/mid_train/step_750.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/post_train/step_1500.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval/post_train/step_4000.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/eval_inputs/flappy_fixed_latency_3.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/hydra/.hydra/hydra.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/hydra/.hydra/overrides.yaml"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/hydra/train_starvla_hydra.log"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_1750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_2750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3250/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3250/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3250/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_3750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_4000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_500/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_500/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_500/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_750/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_750/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/mid_train/step_750/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/post_train/step_4000/_progress/rank_0.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/post_train/step_4000/_progress/rank_1.json"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/latency_bench_eval/post_train/step_4000/flappy/latency_3/episode_metrics.jsonl"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/flappy_fix_latency_3_200ep_7k2steps_stitch/summary.jsonl"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "run": "flappy_fix_latency_3_200ep_7k2steps_stitch",
@@ -412,7 +417,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps",
       "config_name": "flappy_fixed_latency_3_200ep_7k2steps"
     }

latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_4_200ep/artifact.json

@@ -7,13 +7,18 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_4_200ep/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/flappy_fix_latency_4_200ep"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_flappy_200ep",
@@ -64,9 +69,23 @@
         "paper/figures/fig6_latency_robustness/flappy"
       ],
       "paper_identity_confirmed": true,
-      "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
+      "interpretation": "The run and checkpoint path are bound to the paper episode CSV and source manifest. Original episode JSONLs/action logs were not recovered. Training-data source limits are recorded separately.",
       "unpublished_training_data": "The recorded local training directory has no recovered exact HF source. No replacement dataset is assumed."
     }
   ],
+  "evaluation_evidence": {
+    "episodes": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/episodes.csv"
+    },
+    "source_manifest": {
+      "repo": "latency-sensitive-bench/benchmark-datasets",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
+      "path": "evaluations/flappy/latency-transfer-qwenoft/source_manifest.csv"
+    },
+    "checkpoint_run": "flappy_fix_latency_4_200ep",
+    "scope": "Paper run/checkpoint paths and episode-level metrics are retained. Original episode JSONLs and action logs are not recovered."
+  },
   "training_data": []
 }

latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15/artifact.json

@@ -7,19 +7,24 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15/eval/post_train/step_5000.json"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15",
@@ -35,7 +40,7 @@
       "training_data_mix": "flappy_train__bridge",
       "dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost15",
         "source_subdir": null
       },
@@ -81,7 +86,7 @@
       "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
       "training_recipe_dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost15",
         "source_subdir": null
       }
@@ -90,7 +95,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost15"
     }
   ]

latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_exp2/README.md

@@ -4,4 +4,4 @@ Flappy Bird · qwenoft · H1
 
 [Experiment settings and asset manifest](https://huggingface.co/datasets/latency-sensitive-bench/benchmark-datasets/tree/main/experiments/visual-history/archive/flappy)
 
-Download this run at the fixed revision in the task manifest. Keep the checkpoint, original configuration and normalization statistics together. Original local paths describe source provenance. `artifact.json` records the current asset references.
+This historical run shares identical inference weights with another retained run. Download this run’s own config and statistics, and place the canonical checkpoint referenced in `artifact.json` under its local `checkpoints/` directory before native loading. Distinct source run and seed metadata remain preserved. Original local paths describe source provenance. `artifact.json` records the current asset references.

latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_exp2/artifact.json

@@ -7,19 +7,24 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
-      "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_exp2/checkpoints/steps_5000_model.safetensors"
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+      "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_exp2"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_exp2/eval/post_train/step_5000.json"
     }
   ],
-  "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "asset_state": "shared_checkpoint",
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_exp2",
@@ -35,7 +40,7 @@
       "training_data_mix": "flappy_train__bridge",
       "dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost15",
         "source_subdir": null
       },
@@ -81,7 +86,7 @@
       "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
       "training_recipe_dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost15",
         "source_subdir": null
       }
@@ -90,7 +95,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost15"
     }
   ]

latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_seed1/README.md

@@ -4,4 +4,4 @@ Flappy Bird · qwenoft · H1
 
 [Experiment settings and asset manifest](https://huggingface.co/datasets/latency-sensitive-bench/benchmark-datasets/tree/main/experiments/visual-history/archive/flappy)
 
-Download this run at the fixed revision in the task manifest. Keep the checkpoint, original configuration and normalization statistics together. Original local paths describe source provenance. `artifact.json` records the current asset references.
+This historical run shares identical inference weights with another retained run. Download this run’s own config and statistics, and place the canonical checkpoint referenced in `artifact.json` under its local `checkpoints/` directory before native loading. Distinct source run and seed metadata remain preserved. Original local paths describe source provenance. `artifact.json` records the current asset references.

latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_seed1/artifact.json

@@ -7,19 +7,24 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
-      "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_seed1/checkpoints/steps_5000_model.safetensors"
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+      "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15/checkpoints/steps_5000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_seed1"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_seed1/eval/post_train/step_5000.json"
     }
   ],
-  "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "asset_state": "shared_checkpoint",
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost15_seed1",
@@ -35,7 +40,7 @@
       "training_data_mix": "flappy_train__bridge",
       "dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost15",
         "source_subdir": null
       },
@@ -81,7 +86,7 @@
       "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
       "training_recipe_dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost15",
         "source_subdir": null
       }
@@ -90,7 +95,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost15"
     }
   ]

latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost7_exp3/artifact.json

@@ -7,19 +7,24 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost7_exp3/checkpoints/steps_4000_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost7_exp3"
+  },
   "evaluation": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "evaluations/flappy/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost7_exp3/eval/post_train/step_4000.json"
     }
   ],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": true,
+  "paper_checkpoint_binding_confirmed": true,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_bridge_flappy_fixed_latency_3_200ep_7k2steps_ghost7_exp3",
@@ -90,7 +95,7 @@
   "training_data": [
     {
       "repo": "latency-sensitive-bench/benchmark-datasets",
-      "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+      "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
       "path": "latency-aware/flappy/flappy_fixed_latency_3_200ep_7k2steps_ghost7"
     }
   ]

latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_latency_mixed_exp2/artifact.json

@@ -2,12 +2,12 @@
   "condition": "with_latency_prompt",
   "checkpoint": {
     "repo": "latency-sensitive-bench/benchmark-models",
-    "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
     "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_latency_mixed_exp2/checkpoints/model.safetensors"
   },
   "evaluation": {
     "repo": "latency-sensitive-bench/benchmark-datasets",
-    "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+    "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
     "path": "evaluations/flappy/openvla_bridge_flappy_latency_mixed_exp2/eval/post_train/step_5000.json"
   },
   "source": {

latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_latency_mixed_exp2_no_latency_information/artifact.json

@@ -2,12 +2,12 @@
   "condition": "no_latency_information",
   "checkpoint": {
     "repo": "latency-sensitive-bench/benchmark-models",
-    "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
     "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_bridge_flappy_latency_mixed_exp2_no_latency_information/checkpoints/model.safetensors"
   },
   "evaluation": {
     "repo": "latency-sensitive-bench/benchmark-datasets",
-    "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+    "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
     "path": "evaluations/flappy/openvla_bridge_flappy_latency_mixed_exp2_no_latency_information/eval/post_train/step_5000.json"
   },
   "source": {

latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_flappy_fix_latency_1_small/artifact.json

@@ -7,18 +7,23 @@
   "checkpoint": [
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_flappy_fix_latency_1_small/checkpoints/best_model.safetensors"
     },
     {
       "repo": "latency-sensitive-bench/benchmark-models",
-      "revision": "a4099f759f71ffa72d967be3695c8e054cfc377c",
+      "revision": "403deb18146f6780ac28536f78167b21aa18898b",
       "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_flappy_fix_latency_1_small/checkpoints/steps_1500_model.safetensors"
     }
   ],
+  "metadata": {
+    "repo": "latency-sensitive-bench/benchmark-models",
+    "revision": "403deb18146f6780ac28536f78167b21aa18898b",
+    "path": "latency-aware/flappy/vla/starvla-qwenoft-h1/openvla_flappy_fix_latency_1_small"
+  },
   "evaluation": [],
   "asset_state": "complete_bundle",
-  "paper_source_confirmed": false,
+  "paper_checkpoint_binding_confirmed": false,
   "original_training_bindings": [
     {
       "source_repo": "latency-sensitive-bench/openvla_flappy_fix_latency_1_small",
@@ -34,7 +39,7 @@
       "training_data_mix": "flappy_train__bridge",
       "dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/flappy/flappy_fix_latency_1_small",
         "evidence": "Training config uses the exact source dataset name as its local data-root directory. Original metadata and statistics are retained separately."
       },
@@ -78,7 +83,7 @@
       "interpretation": "The paper experiment type is known; an exact plotted checkpoint binding is recorded only when its immutable source manifest establishes it.",
       "training_recipe_dataset_reference": {
         "repo": "latency-sensitive-bench/benchmark-datasets",
-        "revision": "15d17287246bd19940c837d6c8a5327564f969c5",
+        "revision": "27d2fe8badf83302809f5570e3118bd987f15600",
         "path": "latency-aware/flappy/flappy_fix_latency_1_small",
         "evidence": "Training config uses the exact source dataset name as its local data-root directory. Original metadata and statistics are retained separately."
       }