Update CrossSearch to 5,919 retained runs

Index and document all 1,650 added runs: 300 Argon native runs and 1,350 uncapped search comparisons. Preserve the original 4,269-run release and its summary separately.

README.md

@@ -8,19 +8,29 @@ configs:
 ---
 # CrossSearch: 100-task paper set
 
-This resource contains 100 tasks and 4,269 selected, completed evaluation runs: 3,000 native-search runs (10 model configurations, three final attempts per task) and 1,269 search-backend ablation runs (three models, 47 shared tasks, three alternative backends, three attempts). The 600 Opus 5.5/GPT-6.1 Sol runs are included in the 3,000.
+This resource contains **100 tasks and 5,919 retained, completed evaluation runs**. The current evaluation comprises **3,300 native-search runs** (11 model configurations, three final attempts per task) and **1,350 uncapped search-backend runs** (three models, 50 tasks, three alternative backends, three attempts). The previous **1,269 capped search-backend runs** remain available as a separate historical cohort.
+
+## November 2026 update
+
+This update adds **1,650 completed runs**: 300 Gemini 4 Argon native-search runs and 1,350 uncapped Exa, Parallel and Firecrawl runs for GPT-5.6 Sol, Kimi K3 and Gemini 3.8 Flash. Each original task/model/condition/replicate cell is represented once in this new cohort. Earlier incomplete attempts are not substituted for complete results, and valid low and zero scores are retained.
+
+The added files begin with `rollouts-20261104-`. They contain the original final answers, rubric judgments, available model reasoning, saved request/response captures and trace files, with operational credentials and local author/account paths removed. The source archives remain unchanged. Search/fetch call-count limits are disabled in the new search-backend cohort; task time, token, context and response-size limits are unchanged. Keep the uncapped and historical capped cohorts separate when comparing results.
+
 
 ## Files
 
 - tasks.jsonl.gz: expert questions, frozen rubrics and evidence maps.
 - tasks.tar.gz: starting workspace assets, separated from grader-only reference material.
-- rollouts-*.jsonl.gz: saved agent trace files, API request/response captures, final answers, original rewards and criterion judgments. The 669 encrypted source runs are exported as readable derivatives. Source archives remain unchanged; no private key is distributed.
-- run-index.jsonl.gz: a compact lookup of all run IDs, tasks, models, conditions, rewards and the shard holding each trajectory.
+- rollouts-*.jsonl.gz: saved agent trace files, API request/response captures, final answers, original rewards and criterion judgments. The 669 earlier encrypted source runs and all 1,650 added protected archives are exported as readable derivatives. Source archives remain unchanged; no private key is distributed.
+- run-index.jsonl.gz: lookup of all 5,919 retained run IDs, tasks, models, conditions, rewards and trajectory shards.
+- run-index-20261104.jsonl.gz: just the 1,650 runs added in this update, including `release_cohort` and `search_call_limit`.
+- run-index-legacy-4269.jsonl.gz and release-manifest-legacy-4269.json: the previous release indexes, preserved intact.
 - judge-models.csv: one row for each of the 415 audited encrypted extension records, with evidence references and recorded judge labels.
 - CrossSearch-Judge-Prompt.txt and CrossSearch-Agent-Prompt.txt: appendix prompt templates, message roles and source notes.
 - grading-protocol.md: scoring protocol and judge-version uncertainty.
 - judge-model-audit.json: recorded judge fields for the 415 encrypted extension records.
-- analysis-summary.json: frozen paper/blog results.
+- analysis-summary.json and analysis-summary-20261104.json: updated 11-model native results, uncapped search comparisons, efficiency values and reporting definitions. Every metric group is labeled by cohort.
+- analysis-summary-legacy-4269.json: the earlier frozen paper/blog results, retained unchanged.
 - release-manifest.json: final counts, shard checksums, redaction totals and verification status. Check this manifest for completeness.
 
 ## Reading the data
@@ -32,7 +42,7 @@ Read rollout files directly with the standard-library recipe below. The complete
 
 Each rollout retains the original archive hash and source-member hashes. Operational credentials and local author/account paths are removed from publication derivatives. Available reasoning text is retained apart from credential and local-path removals; opaque provider reasoning cannot be converted to visible thoughts. Per-run trace files preserve evaluated prompt wrappers. Native and ablation conditions remain separate; valid low and zero rewards are included.
 
-The evaluation uses three final, complete attempts per task and model/search configuration. Some targets required additional fresh attempts when an earlier attempt did not provide complete answer, reward and rubric-judgment evidence. Those incomplete attempts remain separate from the selected runs in this release. No model was called and no answer was regraded to create this resource.
+The evaluation uses three final, complete attempts per task and model/search configuration within each cohort. Some targets required additional fresh attempts when an earlier attempt did not provide complete answer, reward and rubric-judgment evidence. Those incomplete attempts remain separate from the selected runs in this release. No model was called and no answer was regraded to create this resource.
 
 
 For a direct read without a dataset framework, download a rollout shard and use Python's standard library:
@@ -61,3 +71,13 @@ Some exceptionally large traces use `large_record_parts`. Their normal rollout r
         payload = b''.join(pieces)
         assert hashlib.sha256(payload).hexdigest() == run['large_record_sha256']
         run = json.loads(payload)
+
+## Select the added cohort
+
+    import gzip, json
+    with gzip.open('run-index-20261104.jsonl.gz', 'rt', encoding='utf-8') as f:
+        added = [json.loads(line) for line in f]
+    argon = [r for r in added if r['condition'] == 'native']  # 300
+    uncapped = [r for r in added if r['condition'] != 'native']  # 1,350
+
+The current leaderboard combines the 3,000 earlier native runs with the 300 Argon runs. The current search comparison uses the 1,350 added uncapped runs; it does not pool them with the historical capped runs.

UPLOAD-STATUS.json

@@ -1,20 +1,13 @@
 {
   "status": "complete_private_release",
-  "dataset": "https://huggingface.co/datasets/AfterQuery/CrossSearch",
-  "public_access_verified": false,
-  "gated": false,
   "tasks": 100,
-  "runs": 4269,
-  "native_runs": 3000,
-  "ablation_runs": 1269,
-  "encrypted_source_runs_exported_readably": 669,
-  "all_1312_data_files_committed_sha256_and_size_matched": true,
-  "publisher": "AfterQuery",
-  "publisher_is_anonymous": false,
-  "new_models_or_grades": 0,
-  "documentation_revision": "28fe7fe665cd023e72bda285a1c38d5f404c8aa8",
-  "current_visibility": "private",
-  "private_access_confirmed": true,
-  "owner_instruction": "Keep it private",
-  "historical_unauthenticated_files_checked": 10
-}
+  "retained_runs": 5919,
+  "latest_update_runs": 1650,
+  "current_native_runs": 3300,
+  "current_uncapped_ablation_runs": 1350,
+  "retained_legacy_capped_runs": 1269,
+  "data_commit": "fd44de24985e49be626dc9e7483e14494326fe7e",
+  "visibility": "private",
+  "manifest_sha256": "11ebaa1816f3803bc371bb991f50e1adcd59102b2625f0531ce84423ff612b97",
+  "original_data_files_preserved": true
+}

analysis-summary-20261104.json

@@ -0,0 +1,1858 @@
+{
+  "task_count": 100,
+  "current_native_runs": 3300,
+  "current_uncapped_search_runs": 1350,
+  "retained_legacy_capped_runs": 1269,
+  "retained_run_count": 5919,
+  "added_run_count": 1650,
+  "native_models": [
+    {
+      "id": "google/gemini-4-argon",
+      "name": "Gemini 4 Argon",
+      "lab": "Google",
+      "strict": 86.7,
+      "strict_ci": [
+        80.7,
+        92.3
+      ],
+      "mean": 97.5,
+      "mean_ci": [
+        96.1,
+        98.7
+      ],
+      "all3": 81,
+      "near80": 85.0,
+      "counts": [
+        7,
+        7,
+        5,
+        81
+      ],
+      "hat3": 81.0,
+      "hat3_ci": [
+        73.0,
+        88.0
+      ]
+    },
+    {
+      "id": "anthropic/claude-opus-5.5",
+      "name": "Opus 5.5",
+      "lab": "Anthropic",
+      "strict": 86.0,
+      "strict_ci": [
+        80.0,
+        91.3
+      ],
+      "mean": 97.0,
+      "mean_ci": [
+        95.2,
+        98.4
+      ],
+      "all3": 76,
+      "near80": 81.0,
+      "counts": [
+        6,
+        6,
+        12,
+        76
+      ],
+      "hat3": 76.0,
+      "hat3_ci": [
+        67.0,
+        84.0
+      ]
+    },
+    {
+      "id": "gpt-6-astra",
+      "name": "GPT-6 Astra",
+      "lab": "OpenAI",
+      "strict": 82.0,
+      "strict_ci": [
+        74.7,
+        88.7
+      ],
+      "mean": 95.8,
+      "mean_ci": [
+        93.9,
+        97.5
+      ],
+      "all3": 76,
+      "near80": 63.0,
+      "counts": [
+        13,
+        4,
+        7,
+        76
+      ],
+      "hat3": 76.0,
+      "hat3_ci": [
+        67.0,
+        84.0
+      ]
+    },
+    {
+      "id": "gpt-6.1-sol",
+      "name": "GPT-6.1 Sol",
+      "lab": "OpenAI",
+      "strict": 79.3,
+      "strict_ci": [
+        71.7,
+        86.3
+      ],
+      "mean": 94.5,
+      "mean_ci": [
+        92.3,
+        96.5
+      ],
+      "all3": 73,
+      "near80": 56.5,
+      "counts": [
+        14,
+        7,
+        6,
+        73
+      ],
+      "hat3": 73.0,
+      "hat3_ci": [
+        64.0,
+        82.0
+      ]
+    },
+    {
+      "id": "anthropic/claude-opus-5",
+      "name": "Opus 5",
+      "lab": "Anthropic",
+      "strict": 69.7,
+      "strict_ci": [
+        62.3,
+        77.0
+      ],
+      "mean": 92.2,
+      "mean_ci": [
+        89.7,
+        94.5
+      ],
+      "all3": 51,
+      "near80": 59.3,
+      "counts": [
+        17,
+        8,
+        24,
+        51
+      ],
+      "hat3": 51.0,
+      "hat3_ci": [
+        41.0,
+        60.0
+      ]
+    },
+    {
+      "id": "x-ai/grok-4.6",
+      "name": "Grok 4.6",
+      "lab": "xAI",
+      "strict": 69.0,
+      "strict_ci": [
+        61.3,
+        76.3
+      ],
+      "mean": 91.5,
+      "mean_ci": [
+        88.7,
+        94.0
+      ],
+      "all3": 54,
+      "near80": 57.0,
+      "counts": [
+        17,
+        13,
+        16,
+        54
+      ],
+      "hat3": 54.0,
+      "hat3_ci": [
+        44.0,
+        64.0
+      ]
+    },
+    {
+      "id": "gpt-5.6-sol",
+      "name": "GPT-5.6 Sol",
+      "lab": "OpenAI",
+      "strict": 68.0,
+      "strict_ci": [
+        60.7,
+        75.0
+      ],
+      "mean": 90.7,
+      "mean_ci": [
+        87.5,
+        93.5
+      ],
+      "all3": 47,
+      "near80": 56.2,
+      "counts": [
+        15,
+        13,
+        25,
+        47
+      ],
+      "hat3": 47.0,
+      "hat3_ci": [
+        37.0,
+        57.0
+      ]
+    },
+    {
+      "id": "google/gemini-3.8-flash",
+      "name": "Gemini 3.8 Flash",
+      "lab": "Google",
+      "strict": 58.0,
+      "strict_ci": [
+        50.0,
+        66.0
+      ],
+      "mean": 87.2,
+      "mean_ci": [
+        83.7,
+        90.4
+      ],
+      "all3": 38,
+      "near80": 53.2,
+      "counts": [
+        26,
+        12,
+        24,
+        38
+      ],
+      "hat3": 38.0,
+      "hat3_ci": [
+        29.0,
+        47.0
+      ]
+    },
+    {
+      "id": "moonshotai/kimi-k3",
+      "name": "Kimi K3",
+      "lab": "Moonshot AI",
+      "strict": 41.0,
+      "strict_ci": [
+        33.7,
+        48.7
+      ],
+      "mean": 80.9,
+      "mean_ci": [
+        77.0,
+        84.6
+      ],
+      "all3": 23,
+      "near80": 46.3,
+      "counts": [
+        36,
+        28,
+        13,
+        23
+      ],
+      "hat3": 23.0,
+      "hat3_ci": [
+        15.0,
+        31.0
+      ]
+    },
+    {
+      "id": "anthropic/claude-sonnet-5",
+      "name": "Sonnet 5",
+      "lab": "Anthropic",
+      "strict": 34.3,
+      "strict_ci": [
+        28.0,
+        41.0
+      ],
+      "mean": 74.5,
+      "mean_ci": [
+        69.9,
+        78.9
+      ],
+      "all3": 11,
+      "near80": 39.1,
+      "counts": [
+        39,
+        30,
+        20,
+        11
+      ],
+      "hat3": 11.0,
+      "hat3_ci": [
+        5.0,
+        17.0
+      ]
+    },
+    {
+      "id": "thinkingmachines/inkling",
+      "name": "Inkling",
+      "lab": "Thinking Machines",
+      "strict": 3.0,
+      "strict_ci": [
+        1.3,
+        5.0
+      ],
+      "mean": 41.6,
+      "mean_ci": [
+        37.3,
+        45.9
+      ],
+      "all3": 0,
+      "near80": 7.6,
+      "counts": [
+        91,
+        9,
+        0,
+        0
+      ],
+      "hat3": 0.0,
+      "hat3_ci": [
+        0.0,
+        0.0
+      ]
+    }
+  ],
+  "efficiency": [
+    {
+      "id": "anthropic/claude-opus-5",
+      "name": "Opus 5",
+      "lab": "Anthropic",
+      "strict": 69.7,
+      "spend": 2.57,
+      "cost": 1.43,
+      "minutes": 30.1,
+      "tokens": 74856
+    },
+    {
+      "id": "anthropic/claude-opus-5.5",
+      "name": "Opus 5.5",
+      "lab": "Anthropic",
+      "strict": 86.0,
+      "spend": 1.07,
+      "cost": 0.85,
+      "minutes": 15.5,
+      "tokens": 42376
+    },
+    {
+      "id": "anthropic/claude-sonnet-5",
+      "name": "Sonnet 5",
+      "lab": "Anthropic",
+      "strict": 34.3,
+      "spend": 2.39,
+      "cost": 0.54,
+      "minutes": 17.7,
+      "tokens": 60656
+    },
+    {
+      "id": "google/gemini-3.8-flash",
+      "name": "Gemini 3.8 Flash",
+      "lab": "Google",
+      "strict": 58.0,
+      "spend": 1.93,
+      "cost": 0.85,
+      "minutes": 14.2,
+      "tokens": 55176,
+      "cost_estimated": true
+    },
+    {
+      "id": "gpt-5.6-sol",
+      "name": "GPT-5.6 Sol",
+      "lab": "OpenAI",
+      "strict": 68.0,
+      "spend": 2.32,
+      "cost": 1.21,
+      "minutes": 12.2,
+      "tokens": 30446
+    },
+    {
+      "id": "gpt-6-astra",
+      "name": "GPT-6 Astra",
+      "lab": "OpenAI",
+      "strict": 82.0,
+      "spend": 4.54,
+      "cost": 2.51,
+      "minutes": 10.0,
+      "tokens": 21260
+    },
+    {
+      "id": "gpt-6.1-sol",
+      "name": "GPT-6.1 Sol",
+      "lab": "OpenAI",
+      "strict": 79.3,
+      "spend": 0.84,
+      "cost": 0.48,
+      "minutes": 7.6,
+      "tokens": 22432
+    },
+    {
+      "id": "moonshotai/kimi-k3",
+      "name": "Kimi K3",
+      "lab": "Moonshot AI",
+      "strict": 41.0,
+      "spend": 0.67,
+      "cost": 0.2,
+      "minutes": 14.5,
+      "tokens": 14840,
+      "cost_estimated": true
+    },
+    {
+      "id": "thinkingmachines/inkling",
+      "name": "Inkling",
+      "lab": "Thinking Machines",
+      "strict": 3.0,
+      "spend": 6.55,
+      "cost": 0.13,
+      "minutes": 3.4,
+      "tokens": 12332
+    },
+    {
+      "id": "x-ai/grok-4.6",
+      "name": "Grok 4.6",
+      "lab": "xAI",
+      "strict": 69.0,
+      "spend": 1.17,
+      "cost": 0.76,
+      "minutes": 23.9,
+      "tokens": 59655
+    },
+    {
+      "id": "google/gemini-4-argon",
+      "name": "Gemini 4 Argon",
+      "lab": "Google",
+      "strict": 86.7,
+      "spend": 1.72,
+      "cost": 1.04,
+      "minutes": 15.6,
+      "tokens": 43595,
+      "cost_estimated": true
+    }
+  ],
+  "uncapped_search_comparison": [
+    {
+      "id": "google/gemini-3.8-flash",
+      "name": "Gemini 3.8 Flash",
+      "native": 60.7,
+      "rows": [
+        {
+          "cond": "Exa",
+          "delta": -2.0,
+          "ci": [
+            -11.3,
+            6.7
+          ],
+          "alt": 58.7
+        },
+        {
+          "cond": "Parallel",
+          "delta": 4.0,
+          "ci": [
+            -4.7,
+            12.7
+          ],
+          "alt": 64.7
+        },
+        {
+          "cond": "Firecrawl",
+          "delta": 2.0,
+          "ci": [
+            -6.0,
+            10.0
+          ],
+          "alt": 62.7
+        }
+      ],
+      "abs": [
+        {
+          "cond": "Native",
+          "strict": 60.7,
+          "ci": [
+            50.0,
+            71.3
+          ],
+          "mean": 89.1
+        },
+        {
+          "cond": "Exa",
+          "strict": 58.7,
+          "ci": [
+            46.7,
+            70.7
+          ],
+          "mean": 89.9
+        },
+        {
+          "cond": "Parallel",
+          "strict": 64.7,
+          "ci": [
+            52.7,
+            76.0
+          ],
+          "mean": 91.9
+        },
+        {
+          "cond": "Firecrawl",
+          "strict": 62.7,
+          "ci": [
+            51.3,
+            74.0
+          ],
+          "mean": 91.8
+        }
+      ]
+    },
+    {
+      "id": "gpt-5.6-sol",
+      "name": "GPT-5.6 Sol",
+      "native": 68.7,
+      "rows": [
+        {
+          "cond": "Exa",
+          "delta": 4.0,
+          "ci": [
+            -5.3,
+            12.7
+          ],
+          "alt": 72.7
+        },
+        {
+          "cond": "Parallel",
+          "delta": 2.0,
+          "ci": [
+            -5.3,
+            9.3
+          ],
+          "alt": 70.7
+        },
+        {
+          "cond": "Firecrawl",
+          "delta": 1.3,
+          "ci": [
+            -6.7,
+            9.3
+          ],
+          "alt": 70.0
+        }
+      ],
+      "abs": [
+        {
+          "cond": "Native",
+          "strict": 68.7,
+          "ci": [
+            58.7,
+            78.7
+          ],
+          "mean": 90.5
+        },
+        {
+          "cond": "Exa",
+          "strict": 72.7,
+          "ci": [
+            62.0,
+            82.7
+          ],
+          "mean": 92.6
+        },
+        {
+          "cond": "Parallel",
+          "strict": 70.7,
+          "ci": [
+            59.3,
+            80.7
+          ],
+          "mean": 91.9
+        },
+        {
+          "cond": "Firecrawl",
+          "strict": 70.0,
+          "ci": [
+            59.3,
+            80.0
+          ],
+          "mean": 91.7
+        }
+      ]
+    },
+    {
+      "id": "moonshotai/kimi-k3",
+      "name": "Kimi K3",
+      "native": 41.3,
+      "rows": [
+        {
+          "cond": "Exa",
+          "delta": 12.0,
+          "ci": [
+            2.7,
+            20.7
+          ],
+          "alt": 53.3
+        },
+        {
+          "cond": "Parallel",
+          "delta": 11.3,
+          "ci": [
+            2.0,
+            20.7
+          ],
+          "alt": 52.7
+        },
+        {
+          "cond": "Firecrawl",
+          "delta": 4.7,
+          "ci": [
+            -6.7,
+            16.0
+          ],
+          "alt": 46.0
+        }
+      ],
+      "abs": [
+        {
+          "cond": "Native",
+          "strict": 41.3,
+          "ci": [
+            30.7,
+            52.7
+          ],
+          "mean": 81.0
+        },
+        {
+          "cond": "Exa",
+          "strict": 53.3,
+          "ci": [
+            42.0,
+            64.7
+          ],
+          "mean": 85.7
+        },
+        {
+          "cond": "Parallel",
+          "strict": 52.7,
+          "ci": [
+            42.0,
+            63.3
+          ],
+          "mean": 85.9
+        },
+        {
+          "cond": "Firecrawl",
+          "strict": 46.0,
+          "ci": [
+            34.0,
+            58.0
+          ],
+          "mean": 84.2
+        }
+      ]
+    }
+  ],
+  "failure_categories": {
+    "labels": [
+      "Scope and version",
+      "Retrieval and joining",
+      "Completeness",
+      "Mechanism and configuration",
+      "Threshold and transformation"
+    ],
+    "tasks": [
+      39,
+      22,
+      19,
+      9,
+      11
+    ],
+    "all": {
+      "counts": [
+        483,
+        261,
+        232,
+        146,
+        147
+      ],
+      "total": 1269
+    },
+    "models": [
+      {
+        "id": "google/gemini-4-argon",
+        "name": "Gemini 4 Argon",
+        "counts": [
+          19,
+          6,
+          9,
+          2,
+          4
+        ],
+        "total": 40
+      },
+      {
+        "id": "anthropic/claude-opus-5.5",
+        "name": "Opus 5.5",
+        "counts": [
+          27,
+          4,
+          6,
+          3,
+          2
+        ],
+        "total": 42
+      },
+      {
+        "id": "gpt-6-astra",
+        "name": "GPT-6 Astra",
+        "counts": [
+          13,
+          11,
+          15,
+          6,
+          9
+        ],
+        "total": 54
+      },
+      {
+        "id": "gpt-6.1-sol",
+        "name": "GPT-6.1 Sol",
+        "counts": [
+          18,
+          13,
+          14,
+          8,
+          9
+        ],
+        "total": 62
+      },
+      {
+        "id": "anthropic/claude-opus-5",
+        "name": "Opus 5",
+        "counts": [
+          40,
+          18,
+          8,
+          14,
+          11
+        ],
+        "total": 91
+      },
+      {
+        "id": "x-ai/grok-4.6",
+        "name": "Grok 4.6",
+        "counts": [
+          31,
+          20,
+          14,
+          17,
+          11
+        ],
+        "total": 93
+      },
+      {
+        "id": "gpt-5.6-sol",
+        "name": "GPT-5.6 Sol",
+        "counts": [
+          28,
+          20,
+          23,
+          12,
+          13
+        ],
+        "total": 96
+      },
+      {
+        "id": "google/gemini-3.8-flash",
+        "name": "Gemini 3.8 Flash",
+        "counts": [
+          41,
+          27,
+          27,
+          14,
+          17
+        ],
+        "total": 126
+      },
+      {
+        "id": "moonshotai/kimi-k3",
+        "name": "Kimi K3",
+        "counts": [
+          72,
+          30,
+          32,
+          20,
+          23
+        ],
+        "total": 177
+      },
+      {
+        "id": "anthropic/claude-sonnet-5",
+        "name": "Sonnet 5",
+        "counts": [
+          79,
+          48,
+          30,
+          23,
+          17
+        ],
+        "total": 197
+      },
+      {
+        "id": "thinkingmachines/inkling",
+        "name": "Inkling",
+        "counts": [
+          115,
+          64,
+          54,
+          27,
+          31
+        ],
+        "total": 291
+      }
+    ]
+  },
+  "judge_agreement": [
+    {
+      "key": "gpt-5-6-sol",
+      "name": "GPT-5.6 Sol",
+      "agree_whole": 93.0,
+      "agree_crit": 94.9,
+      "whole": {
+        "fp": 2,
+        "fn": 5,
+        "fails": 96,
+        "passes": 66
+      },
+      "crit": {
+        "fp": 18,
+        "fn": 16,
+        "fails": 199,
+        "passes": 886
+      }
+    },
+    {
+      "key": "opus-5",
+      "name": "Opus 5",
+      "agree_whole": 90.1,
+      "agree_crit": 94.7,
+      "whole": {
+        "fp": 10,
+        "fn": 2,
+        "fails": 96,
+        "passes": 66
+      },
+      "crit": {
+        "fp": 34,
+        "fn": 3,
+        "fails": 199,
+        "passes": 886
+      }
+    },
+    {
+      "key": "gemini-3-7-flash",
+      "name": "Gemini 3.7 Flash",
+      "agree_whole": 89.0,
+      "agree_crit": 93.3,
+      "whole": {
+        "fp": 13,
+        "fn": 1,
+        "fails": 96,
+        "passes": 66
+      },
+      "crit": {
+        "fp": 46,
+        "fn": 6,
+        "fails": 199,
+        "passes": 886
+      }
+    },
+    {
+      "key": "kimi-k3",
+      "name": "Kimi K3",
+      "agree_whole": 84.9,
+      "agree_crit": 92.4,
+      "whole": {
+        "fp": 21,
+        "fn": 0,
+        "fails": 96,
+        "passes": 66
+      },
+      "crit": {
+        "fp": 54,
+        "fn": 8,
+        "fails": 199,
+        "passes": 886
+      }
+    }
+  ],
+  "human_agreement": {
+    "agree_whole": 94.2,
+    "agree_crit": 95.8
+  },
+  "judge_study": {
+    "tasks": 43,
+    "responses": 172
+  },
+  "task_and_rubric_breakdowns": {
+    "task_count": 100,
+    "input_file_count": 435,
+    "file_types": [
+      {
+        "kind": "Image files",
+        "tasks": 25,
+        "files": 55
+      },
+      {
+        "kind": "PDF",
+        "tasks": 66,
+        "files": 138
+      },
+      {
+        "kind": "Audio",
+        "tasks": 9,
+        "files": 9
+      },
+      {
+        "kind": "Video",
+        "tasks": 1,
+        "files": 1
+      },
+      {
+        "kind": "Tables / spreadsheets",
+        "tasks": 60,
+        "files": 88
+      },
+      {
+        "kind": "Text / structured text",
+        "tasks": 45,
+        "files": 144
+      }
+    ],
+    "image_only_pdf_files": 8,
+    "image_only_pdf_tasks": 4,
+    "image_or_image_only_pdf_tasks": 28,
+    "image_audio_video_union_tasks": 33,
+    "nontext_evidence_tasks": 30,
+    "nontext_evidence_rubric_items": 90,
+    "primary_domains": {
+      "engineering": 46,
+      "medicine": 13,
+      "science": 7,
+      "aviation": 1,
+      "law": 7,
+      "public_records": 14,
+      "finance": 9,
+      "other": 3
+    },
+    "rubric_categories": [
+      {
+        "category": "Fact / rule identification",
+        "items": 217,
+        "item_share": 0.3273001508295626,
+        "passed": 6495,
+        "judgements": 7161,
+        "pass_rate": 0.9069962295768748,
+        "models": [
+          {
+            "model": "anthropic/claude-opus-5",
+            "passed": 615,
+            "judgements": 651,
+            "pass_rate": 0.9447004608294931
+          },
+          {
+            "model": "anthropic/claude-opus-5.5",
+            "passed": 645,
+            "judgements": 651,
+            "pass_rate": 0.9907834101382489
+          },
+          {
+            "model": "anthropic/claude-sonnet-5",
+            "passed": 567,
+            "judgements": 651,
+            "pass_rate": 0.8709677419354839
+          },
+          {
+            "model": "google/gemini-3.8-flash",
+            "passed": 596,
+            "judgements": 651,
+            "pass_rate": 0.9155145929339478
+          },
+          {
+            "model": "google/gemini-4-argon",
+            "passed": 636,
+            "judgements": 651,
+            "pass_rate": 0.9769585253456221
+          },
+          {
+            "model": "gpt-5.6-sol",
+            "passed": 627,
+            "judgements": 651,
+            "pass_rate": 0.9631336405529954
+          },
+          {
+            "model": "gpt-6-astra",
+            "passed": 637,
+            "judgements": 651,
+            "pass_rate": 0.978494623655914
+          },
+          {
+            "model": "gpt-6.1-sol",
+            "passed": 632,
+            "judgements": 651,
+            "pass_rate": 0.9708141321044547
+          },
+          {
+            "model": "moonshotai/kimi-k3",
+            "passed": 579,
+            "judgements": 651,
+            "pass_rate": 0.8894009216589862
+          },
+          {
+            "model": "thinkingmachines/inkling",
+            "passed": 355,
+            "judgements": 651,
+            "pass_rate": 0.5453149001536098
+          },
+          {
+            "model": "x-ai/grok-4.6",
+            "passed": 606,
+            "judgements": 651,
+            "pass_rate": 0.9308755760368663
+          }
+        ]
+      },
+      {
+        "category": "Joined application / comparison",
+        "items": 289,
+        "item_share": 0.4358974358974359,
+        "passed": 8107,
+        "judgements": 9537,
+        "pass_rate": 0.8500576701268743,
+        "models": [
+          {
+            "model": "anthropic/claude-opus-5",
+            "passed": 798,
+            "judgements": 867,
+            "pass_rate": 0.9204152249134948
+          },
+          {
+            "model": "anthropic/claude-opus-5.5",
+            "passed": 845,
+            "judgements": 867,
+            "pass_rate": 0.9746251441753172
+          },
+          {
+            "model": "anthropic/claude-sonnet-5",
+            "passed": 629,
+            "judgements": 867,
+            "pass_rate": 0.7254901960784313
+          },
+          {
+            "model": "google/gemini-3.8-flash",
+            "passed": 766,
+            "judgements": 867,
+            "pass_rate": 0.8835063437139562
+          },
+          {
+            "model": "google/gemini-4-argon",
+            "passed": 848,
+            "judgements": 867,
+            "pass_rate": 0.9780853517877739
+          },
+          {
+            "model": "gpt-5.6-sol",
+            "passed": 782,
+            "judgements": 867,
+            "pass_rate": 0.9019607843137255
+          },
+          {
+            "model": "gpt-6-astra",
+            "passed": 836,
+            "judgements": 867,
+            "pass_rate": 0.9642445213379469
+          },
+          {
+            "model": "gpt-6.1-sol",
+            "passed": 819,
+            "judgements": 867,
+            "pass_rate": 0.9446366782006921
+          },
+          {
+            "model": "moonshotai/kimi-k3",
+            "passed": 679,
+            "judgements": 867,
+            "pass_rate": 0.7831603229527105
+          },
+          {
+            "model": "thinkingmachines/inkling",
+            "passed": 313,
+            "judgements": 867,
+            "pass_rate": 0.36101499423298733
+          },
+          {
+            "model": "x-ai/grok-4.6",
+            "passed": 792,
+            "judgements": 867,
+            "pass_rate": 0.9134948096885813
+          }
+        ]
+      },
+      {
+        "category": "Calculation / transformation",
+        "items": 86,
+        "item_share": 0.1297134238310709,
+        "passed": 2445,
+        "judgements": 2838,
+        "pass_rate": 0.861522198731501,
+        "models": [
+          {
+            "model": "anthropic/claude-opus-5",
+            "passed": 238,
+            "judgements": 258,
+            "pass_rate": 0.9224806201550387
+          },
+          {
+            "model": "anthropic/claude-opus-5.5",
+            "passed": 251,
+            "judgements": 258,
+            "pass_rate": 0.9728682170542635
+          },
+          {
+            "model": "anthropic/claude-sonnet-5",
+            "passed": 197,
+            "judgements": 258,
+            "pass_rate": 0.7635658914728682
+          },
+          {
+            "model": "google/gemini-3.8-flash",
+            "passed": 237,
+            "judgements": 258,
+            "pass_rate": 0.9186046511627907
+          },
+          {
+            "model": "google/gemini-4-argon",
+            "passed": 252,
+            "judgements": 258,
+            "pass_rate": 0.9767441860465116
+          },
+          {
+            "model": "gpt-5.6-sol",
+            "passed": 224,
+            "judgements": 258,
+            "pass_rate": 0.8682170542635659
+          },
+          {
+            "model": "gpt-6-astra",
+            "passed": 245,
+            "judgements": 258,
+            "pass_rate": 0.9496124031007752
+          },
+          {
+            "model": "gpt-6.1-sol",
+            "passed": 245,
+            "judgements": 258,
+            "pass_rate": 0.9496124031007752
+          },
+          {
+            "model": "moonshotai/kimi-k3",
+            "passed": 207,
+            "judgements": 258,
+            "pass_rate": 0.8023255813953488
+          },
+          {
+            "model": "thinkingmachines/inkling",
+            "passed": 108,
+            "judgements": 258,
+            "pass_rate": 0.4186046511627907
+          },
+          {
+            "model": "x-ai/grok-4.6",
+            "passed": 241,
+            "judgements": 258,
+            "pass_rate": 0.9341085271317829
+          }
+        ]
+      },
+      {
+        "category": "Grounding / evidence limits",
+        "items": 71,
+        "item_share": 0.10708898944193061,
+        "passed": 2004,
+        "judgements": 2343,
+        "pass_rate": 0.8553137003841229,
+        "models": [
+          {
+            "model": "anthropic/claude-opus-5",
+            "passed": 186,
+            "judgements": 213,
+            "pass_rate": 0.8732394366197183
+          },
+          {
+            "model": "anthropic/claude-opus-5.5",
+            "passed": 196,
+            "judgements": 213,
+            "pass_rate": 0.92018779342723
+          },
+          {
+            "model": "anthropic/claude-sonnet-5",
+            "passed": 166,
+            "judgements": 213,
+            "pass_rate": 0.7793427230046949
+          },
+          {
+            "model": "google/gemini-3.8-flash",
+            "passed": 167,
+            "judgements": 213,
+            "pass_rate": 0.784037558685446
+          },
+          {
+            "model": "google/gemini-4-argon",
+            "passed": 200,
+            "judgements": 213,
+            "pass_rate": 0.9389671361502347
+          },
+          {
+            "model": "gpt-5.6-sol",
+            "passed": 199,
+            "judgements": 213,
+            "pass_rate": 0.9342723004694836
+          },
+          {
+            "model": "gpt-6-astra",
+            "passed": 202,
+            "judgements": 213,
+            "pass_rate": 0.9483568075117371
+          },
+          {
+            "model": "gpt-6.1-sol",
+            "passed": 200,
+            "judgements": 213,
+            "pass_rate": 0.9389671361502347
+          },
+          {
+            "model": "moonshotai/kimi-k3",
+            "passed": 171,
+            "judgements": 213,
+            "pass_rate": 0.8028169014084507
+          },
+          {
+            "model": "thinkingmachines/inkling",
+            "passed": 129,
+            "judgements": 213,
+            "pass_rate": 0.6056338028169014
+          },
+          {
+            "model": "x-ai/grok-4.6",
+            "passed": 188,
+            "judgements": 213,
+            "pass_rate": 0.8826291079812206
+          }
+        ]
+      }
+    ],
+    "category_definitions": {
+      "Fact / rule identification": "Remaining criteria with only one source environment, including direct observations and rule or fact identification.",
+      "Joined application / comparison": "Remaining criteria supported by both workspace and web sources, including comparisons and application of public rules to the local case.",
+      "Calculation / transformation": "Required derived numerical values, arithmetic or numerical methods, time or unit conversions, or numerical feasibility checks.",
+      "Grounding / evidence limits": "Explicit source/locator requirements or requirements to limit claims to what the available evidence establishes."
+    },
+    "category_precedence": [
+      "Calculation / transformation",
+      "Grounding / evidence limits",
+      "Joined application / comparison",
+      "Fact / rule identification"
+    ],
+    "provenance": "Primary domains are original author-provided submission labels. Rubric criterionType is empty for all 663 items. Four reporting categories are exploratory post-hoc coding of primary requirement and original source-support graph, not original labels, a new grading rubric, or a causal failure taxonomy. Calculation and grounding/limits labels were reviewed against criterion text. Remainder splits on whether the linked source graph uses both workspace and web sources.",
+    "rate_definition": "Pooled pass rate of saved per-item boolean judgments across all 11 native-search models, all 100 tasks and all 3 trials per model/task. Each of 663 items has 33 judgments. Categories have different difficulty and composition; rates are descriptive and do not establish causes.",
+    "scoring_basis": "Saved per-item outcomes for non-perfect attempts; validated full-credit records imply every unit-weight item passed. No new grading.",
+    "inventory_definition": "Delivered environment/artifacts files excluding the standard answer_format.md. Original submission URLs without delivered local files are excluded. Image-only PDF means no extractable text and at least one embedded raster image; PDFs with an OCR layer are not identified as image-only."
+  },
+  "cost_note": "Eight model costs are recorded. Gemini 4 Argon, Gemini 3.8 Flash and Kimi K3 costs use saved input, cached-input and output token counts at the November 3, 2026 rate snapshot. Agent cost excludes failed reruns, grading, tools and infrastructure.",
+  "cohort_note": "The 1,269 earlier capped search runs are retained separately and are not pooled with the 1,350 uncapped runs.",
+  "tasks": 100,
+  "runs": 5919,
+  "native_runs": 3300,
+  "ablation_runs": 2619,
+  "uncapped_ablation_runs": 1350,
+  "legacy_capped_ablation_runs": 1269,
+  "current_evaluation_runs": 4650,
+  "metrics": [
+    {
+      "model": "anthropic/claude-opus-5",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.9224380066666666,
+      "full_credit_pass_at_1": 0.6966666666666667,
+      "pass_power_k": {
+        "1": 0.6966666666666667,
+        "2": 0.59,
+        "3": 0.51
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 17,
+        "1": 8,
+        "2": 24,
+        "3": 51
+      }
+    },
+    {
+      "model": "anthropic/claude-opus-5.5",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.96980951,
+      "full_credit_pass_at_1": 0.86,
+      "pass_power_k": {
+        "1": 0.86,
+        "2": 0.8,
+        "3": 0.76
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 6,
+        "1": 6,
+        "2": 12,
+        "3": 76
+      }
+    },
+    {
+      "model": "anthropic/claude-sonnet-5",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.7452871033333334,
+      "full_credit_pass_at_1": 0.3433333333333333,
+      "pass_power_k": {
+        "1": 0.3433333333333333,
+        "2": 0.17666666666666667,
+        "3": 0.11
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 39,
+        "1": 30,
+        "2": 20,
+        "3": 11
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "exa",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8852179432624113,
+      "full_credit_pass_at_1": 0.5460992907801419,
+      "pass_power_k": {
+        "1": 0.5460992907801419,
+        "2": 0.475177304964539,
+        "3": 0.425531914893617
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 17,
+        "1": 3,
+        "2": 7,
+        "3": 20
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "exa",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.89926358,
+      "full_credit_pass_at_1": 0.5866666666666667,
+      "pass_power_k": {
+        "1": 0.5866666666666667,
+        "2": 0.5,
+        "3": 0.46
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 14,
+        "1": 7,
+        "2": 6,
+        "3": 23
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "firecrawl",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8536526382978723,
+      "full_credit_pass_at_1": 0.44680851063829785,
+      "pass_power_k": {
+        "1": 0.44680851063829785,
+        "2": 0.3262411347517731,
+        "3": 0.2765957446808511
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 17,
+        "1": 10,
+        "2": 7,
+        "3": 13
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "firecrawl",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.9175541,
+      "full_credit_pass_at_1": 0.6266666666666667,
+      "pass_power_k": {
+        "1": 0.6266666666666667,
+        "2": 0.5266666666666666,
+        "3": 0.48
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 11,
+        "1": 8,
+        "2": 7,
+        "3": 24
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.8724971966666667,
+      "full_credit_pass_at_1": 0.58,
+      "pass_power_k": {
+        "1": 0.58,
+        "2": 0.46,
+        "3": 0.38
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 26,
+        "1": 12,
+        "2": 24,
+        "3": 38
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "parallel",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8851616595744681,
+      "full_credit_pass_at_1": 0.5673758865248227,
+      "pass_power_k": {
+        "1": 0.5673758865248227,
+        "2": 0.475177304964539,
+        "3": 0.44680851063829785
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 13,
+        "1": 9,
+        "2": 4,
+        "3": 21
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "parallel",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.9194328933333333,
+      "full_credit_pass_at_1": 0.6466666666666666,
+      "pass_power_k": {
+        "1": 0.6466666666666666,
+        "2": 0.56,
+        "3": 0.5
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 12,
+        "1": 4,
+        "2": 9,
+        "3": 25
+      }
+    },
+    {
+      "model": "google/gemini-4-argon",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.975123,
+      "full_credit_pass_at_1": 0.8666666666666667,
+      "pass_power_k": {
+        "1": 0.8666666666666667,
+        "2": 0.8266666666666667,
+        "3": 0.81
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 7,
+        "1": 7,
+        "2": 5,
+        "3": 81
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "exa",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.9034142765957447,
+      "full_credit_pass_at_1": 0.6382978723404256,
+      "pass_power_k": {
+        "1": 0.6382978723404256,
+        "2": 0.5035460992907801,
+        "3": 0.40425531914893614
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 9,
+        "1": 5,
+        "2": 14,
+        "3": 19
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "exa",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.9261211533333333,
+      "full_credit_pass_at_1": 0.7266666666666667,
+      "pass_power_k": {
+        "1": 0.7266666666666667,
+        "2": 0.6466666666666666,
+        "3": 0.6
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 8,
+        "1": 5,
+        "2": 7,
+        "3": 30
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "firecrawl",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8824806028368795,
+      "full_credit_pass_at_1": 0.5957446808510638,
+      "pass_power_k": {
+        "1": 0.5957446808510638,
+        "2": 0.48936170212765956,
+        "3": 0.425531914893617
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 12,
+        "1": 6,
+        "2": 9,
+        "3": 20
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "firecrawl",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.91654498,
+      "full_credit_pass_at_1": 0.7,
+      "pass_power_k": {
+        "1": 0.7,
+        "2": 0.5933333333333334,
+        "3": 0.54
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 7,
+        "1": 8,
+        "2": 8,
+        "3": 27
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.9074536733333334,
+      "full_credit_pass_at_1": 0.68,
+      "pass_power_k": {
+        "1": 0.68,
+        "2": 0.5533333333333333,
+        "3": 0.47
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 15,
+        "1": 13,
+        "2": 25,
+        "3": 47
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "parallel",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8948098156028369,
+      "full_credit_pass_at_1": 0.6099290780141844,
+      "pass_power_k": {
+        "1": 0.6099290780141844,
+        "2": 0.5106382978723404,
+        "3": 0.44680851063829785
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 12,
+        "1": 5,
+        "2": 9,
+        "3": 21
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "parallel",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.9187063466666666,
+      "full_credit_pass_at_1": 0.7066666666666667,
+      "pass_power_k": {
+        "1": 0.7066666666666667,
+        "2": 0.62,
+        "3": 0.58
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 8,
+        "1": 7,
+        "2": 6,
+        "3": 29
+      }
+    },
+    {
+      "model": "gpt-6-astra",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.9581732766666666,
+      "full_credit_pass_at_1": 0.82,
+      "pass_power_k": {
+        "1": 0.82,
+        "2": 0.7833333333333333,
+        "3": 0.76
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 13,
+        "1": 4,
+        "2": 7,
+        "3": 76
+      }
+    },
+    {
+      "model": "gpt-6.1-sol",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.94528174,
+      "full_credit_pass_at_1": 0.7933333333333333,
+      "pass_power_k": {
+        "1": 0.7933333333333333,
+        "2": 0.75,
+        "3": 0.73
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 14,
+        "1": 7,
+        "2": 6,
+        "3": 73
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "exa",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.7981263191489362,
+      "full_credit_pass_at_1": 0.375886524822695,
+      "pass_power_k": {
+        "1": 0.375886524822695,
+        "2": 0.23404255319148937,
+        "3": 0.19148936170212766
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 18,
+        "1": 14,
+        "2": 6,
+        "3": 9
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "exa",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.8570790266666667,
+      "full_credit_pass_at_1": 0.5333333333333333,
+      "pass_power_k": {
+        "1": 0.5333333333333333,
+        "2": 0.41333333333333333,
+        "3": 0.34
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 15,
+        "1": 7,
+        "2": 11,
+        "3": 17
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "firecrawl",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.7683709078014185,
+      "full_credit_pass_at_1": 0.36879432624113473,
+      "pass_power_k": {
+        "1": 0.36879432624113473,
+        "2": 0.2553191489361702,
+        "3": 0.19148936170212766
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 22,
+        "1": 7,
+        "2": 9,
+        "3": 9
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "firecrawl",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.84177482,
+      "full_credit_pass_at_1": 0.46,
+      "pass_power_k": {
+        "1": 0.46,
+        "2": 0.36,
+        "3": 0.3
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 20,
+        "1": 6,
+        "2": 9,
+        "3": 15
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.8092198866666667,
+      "full_credit_pass_at_1": 0.41,
+      "pass_power_k": {
+        "1": 0.41,
+        "2": 0.2733333333333333,
+        "3": 0.23
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 36,
+        "1": 28,
+        "2": 13,
+        "3": 23
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "parallel",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8110883475177305,
+      "full_credit_pass_at_1": 0.375886524822695,
+      "pass_power_k": {
+        "1": 0.375886524822695,
+        "2": 0.2553191489361702,
+        "3": 0.19148936170212766
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 21,
+        "1": 8,
+        "2": 9,
+        "3": 9
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "parallel",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.85915192,
+      "full_credit_pass_at_1": 0.5266666666666666,
+      "pass_power_k": {
+        "1": 0.5266666666666666,
+        "2": 0.3933333333333333,
+        "3": 0.34
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 13,
+        "1": 12,
+        "2": 8,
+        "3": 17
+      }
+    },
+    {
+      "model": "thinkingmachines/inkling",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.41612396,
+      "full_credit_pass_at_1": 0.03,
+      "pass_power_k": {
+        "1": 0.03,
+        "2": 0.0,
+        "3": 0.0
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 91,
+        "1": 9
+      }
+    },
+    {
+      "model": "x-ai/grok-4.6",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.91473413,
+      "full_credit_pass_at_1": 0.69,
+      "pass_power_k": {
+        "1": 0.69,
+        "2": 0.5933333333333334,
+        "3": 0.54
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 17,
+        "1": 13,
+        "2": 16,
+        "3": 54
+      }
+    }
+  ],
+  "no_new_grading_or_model_calls": true,
+  "method": "Metrics are aggregated from retained completed run indexes. Cohort labels distinguish native, current uncapped and historical capped runs; capped and uncapped results are never pooled."
+}

analysis-summary-legacy-4269.json

@@ -0,0 +1 @@
+{"tasks":100,"runs":4269,"native_runs":3000,"ablation_runs":1269,"metrics":[{"model":"anthropic/claude-opus-5","condition":"native","tasks":100,"runs":300,"mean_reward":0.9224380066666666,"full_credit_pass_at_1":0.6966666666666667,"pass_power_k":{"1":0.6966666666666668,"2":0.59,"3":0.51},"perfect_attempts_per_task_histogram":{"2":24,"3":51,"0":17,"1":8}},{"model":"anthropic/claude-opus-5.5","condition":"native","tasks":100,"runs":300,"mean_reward":0.96980951,"full_credit_pass_at_1":0.86,"pass_power_k":{"1":0.86,"2":0.8,"3":0.76},"perfect_attempts_per_task_histogram":{"0":6,"2":12,"3":76,"1":6}},{"model":"anthropic/claude-sonnet-5","condition":"native","tasks":100,"runs":300,"mean_reward":0.7452871033333334,"full_credit_pass_at_1":0.3433333333333333,"pass_power_k":{"1":0.34333333333333327,"2":0.17666666666666667,"3":0.11},"perfect_attempts_per_task_histogram":{"0":39,"1":30,"3":11,"2":20}},{"model":"google/gemini-3.8-flash","condition":"exa","tasks":47,"runs":141,"mean_reward":0.8852179432624113,"full_credit_pass_at_1":0.5460992907801419,"pass_power_k":{"1":0.5460992907801419,"2":0.47517730496453897,"3":0.425531914893617},"perfect_attempts_per_task_histogram":{"0":17,"3":20,"1":3,"2":7}},{"model":"google/gemini-3.8-flash","condition":"firecrawl","tasks":47,"runs":141,"mean_reward":0.8536526382978723,"full_credit_pass_at_1":0.44680851063829785,"pass_power_k":{"1":0.44680851063829785,"2":0.3262411347517731,"3":0.2765957446808511},"perfect_attempts_per_task_histogram":{"1":10,"3":13,"0":17,"2":7}},{"model":"google/gemini-3.8-flash","condition":"native","tasks":100,"runs":300,"mean_reward":0.8724971966666667,"full_credit_pass_at_1":0.58,"pass_power_k":{"1":0.58,"2":0.46,"3":0.38},"perfect_attempts_per_task_histogram":{"3":38,"2":24,"0":26,"1":12}},{"model":"google/gemini-3.8-flash","condition":"parallel","tasks":47,"runs":141,"mean_reward":0.8851616595744681,"full_credit_pass_at_1":0.5673758865248227,"pass_power_k":{"1":0.5673758865248227,"2":0.47517730496453897,"3":0.44680851063829785},"perfect_attempts_per_task_histogram":{"3":21,"1":9,"2":4,"0":13}},{"model":"gpt-5.6-sol","condition":"exa","tasks":47,"runs":141,"mean_reward":0.9034142765957447,"full_credit_pass_at_1":0.6382978723404256,"pass_power_k":{"1":0.6382978723404256,"2":0.5035460992907802,"3":0.40425531914893614},"perfect_attempts_per_task_histogram":{"0":9,"2":14,"1":5,"3":19}},{"model":"gpt-5.6-sol","condition":"firecrawl","tasks":47,"runs":141,"mean_reward":0.8824806028368795,"full_credit_pass_at_1":0.5957446808510638,"pass_power_k":{"1":0.5957446808510638,"2":0.48936170212765956,"3":0.425531914893617},"perfect_attempts_per_task_histogram":{"2":9,"3":20,"0":12,"1":6}},{"model":"gpt-5.6-sol","condition":"native","tasks":100,"runs":300,"mean_reward":0.9074536733333334,"full_credit_pass_at_1":0.68,"pass_power_k":{"1":0.68,"2":0.5533333333333333,"3":0.47},"perfect_attempts_per_task_histogram":{"1":13,"0":15,"2":25,"3":47}},{"model":"gpt-5.6-sol","condition":"parallel","tasks":47,"runs":141,"mean_reward":0.8948098156028369,"full_credit_pass_at_1":0.6099290780141844,"pass_power_k":{"1":0.6099290780141844,"2":0.5106382978723404,"3":0.44680851063829785},"perfect_attempts_per_task_histogram":{"2":9,"0":12,"3":21,"1":5}},{"model":"gpt-6-astra","condition":"native","tasks":100,"runs":300,"mean_reward":0.9581732766666666,"full_credit_pass_at_1":0.82,"pass_power_k":{"1":0.82,"2":0.7833333333333333,"3":0.76},"perfect_attempts_per_task_histogram":{"3":76,"0":13,"2":7,"1":4}},{"model":"gpt-6.1-sol","condition":"native","tasks":100,"runs":300,"mean_reward":0.94528174,"full_credit_pass_at_1":0.7933333333333333,"pass_power_k":{"1":0.7933333333333333,"2":0.75,"3":0.73},"perfect_attempts_per_task_histogram":{"3":73,"0":14,"1":7,"2":6}},{"model":"moonshotai/kimi-k3","condition":"exa","tasks":47,"runs":141,"mean_reward":0.7981263191489362,"full_credit_pass_at_1":0.375886524822695,"pass_power_k":{"1":0.3758865248226951,"2":0.23404255319148937,"3":0.19148936170212766},"perfect_attempts_per_task_histogram":{"3":9,"0":18,"1":14,"2":6}},{"model":"moonshotai/kimi-k3","condition":"firecrawl","tasks":47,"runs":141,"mean_reward":0.7683709078014185,"full_credit_pass_at_1":0.36879432624113473,"pass_power_k":{"1":0.36879432624113473,"2":0.2553191489361702,"3":0.19148936170212766},"perfect_attempts_per_task_histogram":{"0":22,"1":7,"2":9,"3":9}},{"model":"moonshotai/kimi-k3","condition":"native","tasks":100,"runs":300,"mean_reward":0.8092198866666667,"full_credit_pass_at_1":0.41,"pass_power_k":{"1":0.41,"2":0.2733333333333333,"3":0.23},"perfect_attempts_per_task_histogram":{"1":28,"0":36,"3":23,"2":13}},{"model":"moonshotai/kimi-k3","condition":"parallel","tasks":47,"runs":141,"mean_reward":0.8110883475177305,"full_credit_pass_at_1":0.375886524822695,"pass_power_k":{"1":0.3758865248226951,"2":0.2553191489361702,"3":0.19148936170212766},"perfect_attempts_per_task_histogram":{"1":8,"0":21,"2":9,"3":9}},{"model":"thinkingmachines/inkling","condition":"native","tasks":100,"runs":300,"mean_reward":0.41612396,"full_credit_pass_at_1":0.03,"pass_power_k":{"1":0.03,"2":0.0,"3":0.0},"perfect_attempts_per_task_histogram":{"0":91,"1":9}},{"model":"x-ai/grok-4.6","condition":"native","tasks":100,"runs":300,"mean_reward":0.9147341299999999,"full_credit_pass_at_1":0.69,"pass_power_k":{"1":0.69,"2":0.5933333333333334,"3":0.54},"perfect_attempts_per_task_histogram":{"3":54,"1":13,"2":16,"0":17}}],"no_new_grading_or_model_calls":true,"method":"Mean reward and full-credit probability from three selected final attempts per task/configuration; pass^k is the mean probability that all k attempts pass, comb(c,k)/comb(3,k)."}

analysis-summary.json

@@ -1 +1,1858 @@
-{"tasks":100,"runs":4269,"native_runs":3000,"ablation_runs":1269,"metrics":[{"model":"anthropic/claude-opus-5","condition":"native","tasks":100,"runs":300,"mean_reward":0.9224380066666666,"full_credit_pass_at_1":0.6966666666666667,"pass_power_k":{"1":0.6966666666666668,"2":0.59,"3":0.51},"perfect_attempts_per_task_histogram":{"2":24,"3":51,"0":17,"1":8}},{"model":"anthropic/claude-opus-5.5","condition":"native","tasks":100,"runs":300,"mean_reward":0.96980951,"full_credit_pass_at_1":0.86,"pass_power_k":{"1":0.86,"2":0.8,"3":0.76},"perfect_attempts_per_task_histogram":{"0":6,"2":12,"3":76,"1":6}},{"model":"anthropic/claude-sonnet-5","condition":"native","tasks":100,"runs":300,"mean_reward":0.7452871033333334,"full_credit_pass_at_1":0.3433333333333333,"pass_power_k":{"1":0.34333333333333327,"2":0.17666666666666667,"3":0.11},"perfect_attempts_per_task_histogram":{"0":39,"1":30,"3":11,"2":20}},{"model":"google/gemini-3.8-flash","condition":"exa","tasks":47,"runs":141,"mean_reward":0.8852179432624113,"full_credit_pass_at_1":0.5460992907801419,"pass_power_k":{"1":0.5460992907801419,"2":0.47517730496453897,"3":0.425531914893617},"perfect_attempts_per_task_histogram":{"0":17,"3":20,"1":3,"2":7}},{"model":"google/gemini-3.8-flash","condition":"firecrawl","tasks":47,"runs":141,"mean_reward":0.8536526382978723,"full_credit_pass_at_1":0.44680851063829785,"pass_power_k":{"1":0.44680851063829785,"2":0.3262411347517731,"3":0.2765957446808511},"perfect_attempts_per_task_histogram":{"1":10,"3":13,"0":17,"2":7}},{"model":"google/gemini-3.8-flash","condition":"native","tasks":100,"runs":300,"mean_reward":0.8724971966666667,"full_credit_pass_at_1":0.58,"pass_power_k":{"1":0.58,"2":0.46,"3":0.38},"perfect_attempts_per_task_histogram":{"3":38,"2":24,"0":26,"1":12}},{"model":"google/gemini-3.8-flash","condition":"parallel","tasks":47,"runs":141,"mean_reward":0.8851616595744681,"full_credit_pass_at_1":0.5673758865248227,"pass_power_k":{"1":0.5673758865248227,"2":0.47517730496453897,"3":0.44680851063829785},"perfect_attempts_per_task_histogram":{"3":21,"1":9,"2":4,"0":13}},{"model":"gpt-5.6-sol","condition":"exa","tasks":47,"runs":141,"mean_reward":0.9034142765957447,"full_credit_pass_at_1":0.6382978723404256,"pass_power_k":{"1":0.6382978723404256,"2":0.5035460992907802,"3":0.40425531914893614},"perfect_attempts_per_task_histogram":{"0":9,"2":14,"1":5,"3":19}},{"model":"gpt-5.6-sol","condition":"firecrawl","tasks":47,"runs":141,"mean_reward":0.8824806028368795,"full_credit_pass_at_1":0.5957446808510638,"pass_power_k":{"1":0.5957446808510638,"2":0.48936170212765956,"3":0.425531914893617},"perfect_attempts_per_task_histogram":{"2":9,"3":20,"0":12,"1":6}},{"model":"gpt-5.6-sol","condition":"native","tasks":100,"runs":300,"mean_reward":0.9074536733333334,"full_credit_pass_at_1":0.68,"pass_power_k":{"1":0.68,"2":0.5533333333333333,"3":0.47},"perfect_attempts_per_task_histogram":{"1":13,"0":15,"2":25,"3":47}},{"model":"gpt-5.6-sol","condition":"parallel","tasks":47,"runs":141,"mean_reward":0.8948098156028369,"full_credit_pass_at_1":0.6099290780141844,"pass_power_k":{"1":0.6099290780141844,"2":0.5106382978723404,"3":0.44680851063829785},"perfect_attempts_per_task_histogram":{"2":9,"0":12,"3":21,"1":5}},{"model":"gpt-6-astra","condition":"native","tasks":100,"runs":300,"mean_reward":0.9581732766666666,"full_credit_pass_at_1":0.82,"pass_power_k":{"1":0.82,"2":0.7833333333333333,"3":0.76},"perfect_attempts_per_task_histogram":{"3":76,"0":13,"2":7,"1":4}},{"model":"gpt-6.1-sol","condition":"native","tasks":100,"runs":300,"mean_reward":0.94528174,"full_credit_pass_at_1":0.7933333333333333,"pass_power_k":{"1":0.7933333333333333,"2":0.75,"3":0.73},"perfect_attempts_per_task_histogram":{"3":73,"0":14,"1":7,"2":6}},{"model":"moonshotai/kimi-k3","condition":"exa","tasks":47,"runs":141,"mean_reward":0.7981263191489362,"full_credit_pass_at_1":0.375886524822695,"pass_power_k":{"1":0.3758865248226951,"2":0.23404255319148937,"3":0.19148936170212766},"perfect_attempts_per_task_histogram":{"3":9,"0":18,"1":14,"2":6}},{"model":"moonshotai/kimi-k3","condition":"firecrawl","tasks":47,"runs":141,"mean_reward":0.7683709078014185,"full_credit_pass_at_1":0.36879432624113473,"pass_power_k":{"1":0.36879432624113473,"2":0.2553191489361702,"3":0.19148936170212766},"perfect_attempts_per_task_histogram":{"0":22,"1":7,"2":9,"3":9}},{"model":"moonshotai/kimi-k3","condition":"native","tasks":100,"runs":300,"mean_reward":0.8092198866666667,"full_credit_pass_at_1":0.41,"pass_power_k":{"1":0.41,"2":0.2733333333333333,"3":0.23},"perfect_attempts_per_task_histogram":{"1":28,"0":36,"3":23,"2":13}},{"model":"moonshotai/kimi-k3","condition":"parallel","tasks":47,"runs":141,"mean_reward":0.8110883475177305,"full_credit_pass_at_1":0.375886524822695,"pass_power_k":{"1":0.3758865248226951,"2":0.2553191489361702,"3":0.19148936170212766},"perfect_attempts_per_task_histogram":{"1":8,"0":21,"2":9,"3":9}},{"model":"thinkingmachines/inkling","condition":"native","tasks":100,"runs":300,"mean_reward":0.41612396,"full_credit_pass_at_1":0.03,"pass_power_k":{"1":0.03,"2":0.0,"3":0.0},"perfect_attempts_per_task_histogram":{"0":91,"1":9}},{"model":"x-ai/grok-4.6","condition":"native","tasks":100,"runs":300,"mean_reward":0.9147341299999999,"full_credit_pass_at_1":0.69,"pass_power_k":{"1":0.69,"2":0.5933333333333334,"3":0.54},"perfect_attempts_per_task_histogram":{"3":54,"1":13,"2":16,"0":17}}],"no_new_grading_or_model_calls":true,"method":"Mean reward and full-credit probability from three selected final attempts per task/configuration; pass^k is the mean probability that all k attempts pass, comb(c,k)/comb(3,k)."}
+{
+  "task_count": 100,
+  "current_native_runs": 3300,
+  "current_uncapped_search_runs": 1350,
+  "retained_legacy_capped_runs": 1269,
+  "retained_run_count": 5919,
+  "added_run_count": 1650,
+  "native_models": [
+    {
+      "id": "google/gemini-4-argon",
+      "name": "Gemini 4 Argon",
+      "lab": "Google",
+      "strict": 86.7,
+      "strict_ci": [
+        80.7,
+        92.3
+      ],
+      "mean": 97.5,
+      "mean_ci": [
+        96.1,
+        98.7
+      ],
+      "all3": 81,
+      "near80": 85.0,
+      "counts": [
+        7,
+        7,
+        5,
+        81
+      ],
+      "hat3": 81.0,
+      "hat3_ci": [
+        73.0,
+        88.0
+      ]
+    },
+    {
+      "id": "anthropic/claude-opus-5.5",
+      "name": "Opus 5.5",
+      "lab": "Anthropic",
+      "strict": 86.0,
+      "strict_ci": [
+        80.0,
+        91.3
+      ],
+      "mean": 97.0,
+      "mean_ci": [
+        95.2,
+        98.4
+      ],
+      "all3": 76,
+      "near80": 81.0,
+      "counts": [
+        6,
+        6,
+        12,
+        76
+      ],
+      "hat3": 76.0,
+      "hat3_ci": [
+        67.0,
+        84.0
+      ]
+    },
+    {
+      "id": "gpt-6-astra",
+      "name": "GPT-6 Astra",
+      "lab": "OpenAI",
+      "strict": 82.0,
+      "strict_ci": [
+        74.7,
+        88.7
+      ],
+      "mean": 95.8,
+      "mean_ci": [
+        93.9,
+        97.5
+      ],
+      "all3": 76,
+      "near80": 63.0,
+      "counts": [
+        13,
+        4,
+        7,
+        76
+      ],
+      "hat3": 76.0,
+      "hat3_ci": [
+        67.0,
+        84.0
+      ]
+    },
+    {
+      "id": "gpt-6.1-sol",
+      "name": "GPT-6.1 Sol",
+      "lab": "OpenAI",
+      "strict": 79.3,
+      "strict_ci": [
+        71.7,
+        86.3
+      ],
+      "mean": 94.5,
+      "mean_ci": [
+        92.3,
+        96.5
+      ],
+      "all3": 73,
+      "near80": 56.5,
+      "counts": [
+        14,
+        7,
+        6,
+        73
+      ],
+      "hat3": 73.0,
+      "hat3_ci": [
+        64.0,
+        82.0
+      ]
+    },
+    {
+      "id": "anthropic/claude-opus-5",
+      "name": "Opus 5",
+      "lab": "Anthropic",
+      "strict": 69.7,
+      "strict_ci": [
+        62.3,
+        77.0
+      ],
+      "mean": 92.2,
+      "mean_ci": [
+        89.7,
+        94.5
+      ],
+      "all3": 51,
+      "near80": 59.3,
+      "counts": [
+        17,
+        8,
+        24,
+        51
+      ],
+      "hat3": 51.0,
+      "hat3_ci": [
+        41.0,
+        60.0
+      ]
+    },
+    {
+      "id": "x-ai/grok-4.6",
+      "name": "Grok 4.6",
+      "lab": "xAI",
+      "strict": 69.0,
+      "strict_ci": [
+        61.3,
+        76.3
+      ],
+      "mean": 91.5,
+      "mean_ci": [
+        88.7,
+        94.0
+      ],
+      "all3": 54,
+      "near80": 57.0,
+      "counts": [
+        17,
+        13,
+        16,
+        54
+      ],
+      "hat3": 54.0,
+      "hat3_ci": [
+        44.0,
+        64.0
+      ]
+    },
+    {
+      "id": "gpt-5.6-sol",
+      "name": "GPT-5.6 Sol",
+      "lab": "OpenAI",
+      "strict": 68.0,
+      "strict_ci": [
+        60.7,
+        75.0
+      ],
+      "mean": 90.7,
+      "mean_ci": [
+        87.5,
+        93.5
+      ],
+      "all3": 47,
+      "near80": 56.2,
+      "counts": [
+        15,
+        13,
+        25,
+        47
+      ],
+      "hat3": 47.0,
+      "hat3_ci": [
+        37.0,
+        57.0
+      ]
+    },
+    {
+      "id": "google/gemini-3.8-flash",
+      "name": "Gemini 3.8 Flash",
+      "lab": "Google",
+      "strict": 58.0,
+      "strict_ci": [
+        50.0,
+        66.0
+      ],
+      "mean": 87.2,
+      "mean_ci": [
+        83.7,
+        90.4
+      ],
+      "all3": 38,
+      "near80": 53.2,
+      "counts": [
+        26,
+        12,
+        24,
+        38
+      ],
+      "hat3": 38.0,
+      "hat3_ci": [
+        29.0,
+        47.0
+      ]
+    },
+    {
+      "id": "moonshotai/kimi-k3",
+      "name": "Kimi K3",
+      "lab": "Moonshot AI",
+      "strict": 41.0,
+      "strict_ci": [
+        33.7,
+        48.7
+      ],
+      "mean": 80.9,
+      "mean_ci": [
+        77.0,
+        84.6
+      ],
+      "all3": 23,
+      "near80": 46.3,
+      "counts": [
+        36,
+        28,
+        13,
+        23
+      ],
+      "hat3": 23.0,
+      "hat3_ci": [
+        15.0,
+        31.0
+      ]
+    },
+    {
+      "id": "anthropic/claude-sonnet-5",
+      "name": "Sonnet 5",
+      "lab": "Anthropic",
+      "strict": 34.3,
+      "strict_ci": [
+        28.0,
+        41.0
+      ],
+      "mean": 74.5,
+      "mean_ci": [
+        69.9,
+        78.9
+      ],
+      "all3": 11,
+      "near80": 39.1,
+      "counts": [
+        39,
+        30,
+        20,
+        11
+      ],
+      "hat3": 11.0,
+      "hat3_ci": [
+        5.0,
+        17.0
+      ]
+    },
+    {
+      "id": "thinkingmachines/inkling",
+      "name": "Inkling",
+      "lab": "Thinking Machines",
+      "strict": 3.0,
+      "strict_ci": [
+        1.3,
+        5.0
+      ],
+      "mean": 41.6,
+      "mean_ci": [
+        37.3,
+        45.9
+      ],
+      "all3": 0,
+      "near80": 7.6,
+      "counts": [
+        91,
+        9,
+        0,
+        0
+      ],
+      "hat3": 0.0,
+      "hat3_ci": [
+        0.0,
+        0.0
+      ]
+    }
+  ],
+  "efficiency": [
+    {
+      "id": "anthropic/claude-opus-5",
+      "name": "Opus 5",
+      "lab": "Anthropic",
+      "strict": 69.7,
+      "spend": 2.57,
+      "cost": 1.43,
+      "minutes": 30.1,
+      "tokens": 74856
+    },
+    {
+      "id": "anthropic/claude-opus-5.5",
+      "name": "Opus 5.5",
+      "lab": "Anthropic",
+      "strict": 86.0,
+      "spend": 1.07,
+      "cost": 0.85,
+      "minutes": 15.5,
+      "tokens": 42376
+    },
+    {
+      "id": "anthropic/claude-sonnet-5",
+      "name": "Sonnet 5",
+      "lab": "Anthropic",
+      "strict": 34.3,
+      "spend": 2.39,
+      "cost": 0.54,
+      "minutes": 17.7,
+      "tokens": 60656
+    },
+    {
+      "id": "google/gemini-3.8-flash",
+      "name": "Gemini 3.8 Flash",
+      "lab": "Google",
+      "strict": 58.0,
+      "spend": 1.93,
+      "cost": 0.85,
+      "minutes": 14.2,
+      "tokens": 55176,
+      "cost_estimated": true
+    },
+    {
+      "id": "gpt-5.6-sol",
+      "name": "GPT-5.6 Sol",
+      "lab": "OpenAI",
+      "strict": 68.0,
+      "spend": 2.32,
+      "cost": 1.21,
+      "minutes": 12.2,
+      "tokens": 30446
+    },
+    {
+      "id": "gpt-6-astra",
+      "name": "GPT-6 Astra",
+      "lab": "OpenAI",
+      "strict": 82.0,
+      "spend": 4.54,
+      "cost": 2.51,
+      "minutes": 10.0,
+      "tokens": 21260
+    },
+    {
+      "id": "gpt-6.1-sol",
+      "name": "GPT-6.1 Sol",
+      "lab": "OpenAI",
+      "strict": 79.3,
+      "spend": 0.84,
+      "cost": 0.48,
+      "minutes": 7.6,
+      "tokens": 22432
+    },
+    {
+      "id": "moonshotai/kimi-k3",
+      "name": "Kimi K3",
+      "lab": "Moonshot AI",
+      "strict": 41.0,
+      "spend": 0.67,
+      "cost": 0.2,
+      "minutes": 14.5,
+      "tokens": 14840,
+      "cost_estimated": true
+    },
+    {
+      "id": "thinkingmachines/inkling",
+      "name": "Inkling",
+      "lab": "Thinking Machines",
+      "strict": 3.0,
+      "spend": 6.55,
+      "cost": 0.13,
+      "minutes": 3.4,
+      "tokens": 12332
+    },
+    {
+      "id": "x-ai/grok-4.6",
+      "name": "Grok 4.6",
+      "lab": "xAI",
+      "strict": 69.0,
+      "spend": 1.17,
+      "cost": 0.76,
+      "minutes": 23.9,
+      "tokens": 59655
+    },
+    {
+      "id": "google/gemini-4-argon",
+      "name": "Gemini 4 Argon",
+      "lab": "Google",
+      "strict": 86.7,
+      "spend": 1.72,
+      "cost": 1.04,
+      "minutes": 15.6,
+      "tokens": 43595,
+      "cost_estimated": true
+    }
+  ],
+  "uncapped_search_comparison": [
+    {
+      "id": "google/gemini-3.8-flash",
+      "name": "Gemini 3.8 Flash",
+      "native": 60.7,
+      "rows": [
+        {
+          "cond": "Exa",
+          "delta": -2.0,
+          "ci": [
+            -11.3,
+            6.7
+          ],
+          "alt": 58.7
+        },
+        {
+          "cond": "Parallel",
+          "delta": 4.0,
+          "ci": [
+            -4.7,
+            12.7
+          ],
+          "alt": 64.7
+        },
+        {
+          "cond": "Firecrawl",
+          "delta": 2.0,
+          "ci": [
+            -6.0,
+            10.0
+          ],
+          "alt": 62.7
+        }
+      ],
+      "abs": [
+        {
+          "cond": "Native",
+          "strict": 60.7,
+          "ci": [
+            50.0,
+            71.3
+          ],
+          "mean": 89.1
+        },
+        {
+          "cond": "Exa",
+          "strict": 58.7,
+          "ci": [
+            46.7,
+            70.7
+          ],
+          "mean": 89.9
+        },
+        {
+          "cond": "Parallel",
+          "strict": 64.7,
+          "ci": [
+            52.7,
+            76.0
+          ],
+          "mean": 91.9
+        },
+        {
+          "cond": "Firecrawl",
+          "strict": 62.7,
+          "ci": [
+            51.3,
+            74.0
+          ],
+          "mean": 91.8
+        }
+      ]
+    },
+    {
+      "id": "gpt-5.6-sol",
+      "name": "GPT-5.6 Sol",
+      "native": 68.7,
+      "rows": [
+        {
+          "cond": "Exa",
+          "delta": 4.0,
+          "ci": [
+            -5.3,
+            12.7
+          ],
+          "alt": 72.7
+        },
+        {
+          "cond": "Parallel",
+          "delta": 2.0,
+          "ci": [
+            -5.3,
+            9.3
+          ],
+          "alt": 70.7
+        },
+        {
+          "cond": "Firecrawl",
+          "delta": 1.3,
+          "ci": [
+            -6.7,
+            9.3
+          ],
+          "alt": 70.0
+        }
+      ],
+      "abs": [
+        {
+          "cond": "Native",
+          "strict": 68.7,
+          "ci": [
+            58.7,
+            78.7
+          ],
+          "mean": 90.5
+        },
+        {
+          "cond": "Exa",
+          "strict": 72.7,
+          "ci": [
+            62.0,
+            82.7
+          ],
+          "mean": 92.6
+        },
+        {
+          "cond": "Parallel",
+          "strict": 70.7,
+          "ci": [
+            59.3,
+            80.7
+          ],
+          "mean": 91.9
+        },
+        {
+          "cond": "Firecrawl",
+          "strict": 70.0,
+          "ci": [
+            59.3,
+            80.0
+          ],
+          "mean": 91.7
+        }
+      ]
+    },
+    {
+      "id": "moonshotai/kimi-k3",
+      "name": "Kimi K3",
+      "native": 41.3,
+      "rows": [
+        {
+          "cond": "Exa",
+          "delta": 12.0,
+          "ci": [
+            2.7,
+            20.7
+          ],
+          "alt": 53.3
+        },
+        {
+          "cond": "Parallel",
+          "delta": 11.3,
+          "ci": [
+            2.0,
+            20.7
+          ],
+          "alt": 52.7
+        },
+        {
+          "cond": "Firecrawl",
+          "delta": 4.7,
+          "ci": [
+            -6.7,
+            16.0
+          ],
+          "alt": 46.0
+        }
+      ],
+      "abs": [
+        {
+          "cond": "Native",
+          "strict": 41.3,
+          "ci": [
+            30.7,
+            52.7
+          ],
+          "mean": 81.0
+        },
+        {
+          "cond": "Exa",
+          "strict": 53.3,
+          "ci": [
+            42.0,
+            64.7
+          ],
+          "mean": 85.7
+        },
+        {
+          "cond": "Parallel",
+          "strict": 52.7,
+          "ci": [
+            42.0,
+            63.3
+          ],
+          "mean": 85.9
+        },
+        {
+          "cond": "Firecrawl",
+          "strict": 46.0,
+          "ci": [
+            34.0,
+            58.0
+          ],
+          "mean": 84.2
+        }
+      ]
+    }
+  ],
+  "failure_categories": {
+    "labels": [
+      "Scope and version",
+      "Retrieval and joining",
+      "Completeness",
+      "Mechanism and configuration",
+      "Threshold and transformation"
+    ],
+    "tasks": [
+      39,
+      22,
+      19,
+      9,
+      11
+    ],
+    "all": {
+      "counts": [
+        483,
+        261,
+        232,
+        146,
+        147
+      ],
+      "total": 1269
+    },
+    "models": [
+      {
+        "id": "google/gemini-4-argon",
+        "name": "Gemini 4 Argon",
+        "counts": [
+          19,
+          6,
+          9,
+          2,
+          4
+        ],
+        "total": 40
+      },
+      {
+        "id": "anthropic/claude-opus-5.5",
+        "name": "Opus 5.5",
+        "counts": [
+          27,
+          4,
+          6,
+          3,
+          2
+        ],
+        "total": 42
+      },
+      {
+        "id": "gpt-6-astra",
+        "name": "GPT-6 Astra",
+        "counts": [
+          13,
+          11,
+          15,
+          6,
+          9
+        ],
+        "total": 54
+      },
+      {
+        "id": "gpt-6.1-sol",
+        "name": "GPT-6.1 Sol",
+        "counts": [
+          18,
+          13,
+          14,
+          8,
+          9
+        ],
+        "total": 62
+      },
+      {
+        "id": "anthropic/claude-opus-5",
+        "name": "Opus 5",
+        "counts": [
+          40,
+          18,
+          8,
+          14,
+          11
+        ],
+        "total": 91
+      },
+      {
+        "id": "x-ai/grok-4.6",
+        "name": "Grok 4.6",
+        "counts": [
+          31,
+          20,
+          14,
+          17,
+          11
+        ],
+        "total": 93
+      },
+      {
+        "id": "gpt-5.6-sol",
+        "name": "GPT-5.6 Sol",
+        "counts": [
+          28,
+          20,
+          23,
+          12,
+          13
+        ],
+        "total": 96
+      },
+      {
+        "id": "google/gemini-3.8-flash",
+        "name": "Gemini 3.8 Flash",
+        "counts": [
+          41,
+          27,
+          27,
+          14,
+          17
+        ],
+        "total": 126
+      },
+      {
+        "id": "moonshotai/kimi-k3",
+        "name": "Kimi K3",
+        "counts": [
+          72,
+          30,
+          32,
+          20,
+          23
+        ],
+        "total": 177
+      },
+      {
+        "id": "anthropic/claude-sonnet-5",
+        "name": "Sonnet 5",
+        "counts": [
+          79,
+          48,
+          30,
+          23,
+          17
+        ],
+        "total": 197
+      },
+      {
+        "id": "thinkingmachines/inkling",
+        "name": "Inkling",
+        "counts": [
+          115,
+          64,
+          54,
+          27,
+          31
+        ],
+        "total": 291
+      }
+    ]
+  },
+  "judge_agreement": [
+    {
+      "key": "gpt-5-6-sol",
+      "name": "GPT-5.6 Sol",
+      "agree_whole": 93.0,
+      "agree_crit": 94.9,
+      "whole": {
+        "fp": 2,
+        "fn": 5,
+        "fails": 96,
+        "passes": 66
+      },
+      "crit": {
+        "fp": 18,
+        "fn": 16,
+        "fails": 199,
+        "passes": 886
+      }
+    },
+    {
+      "key": "opus-5",
+      "name": "Opus 5",
+      "agree_whole": 90.1,
+      "agree_crit": 94.7,
+      "whole": {
+        "fp": 10,
+        "fn": 2,
+        "fails": 96,
+        "passes": 66
+      },
+      "crit": {
+        "fp": 34,
+        "fn": 3,
+        "fails": 199,
+        "passes": 886
+      }
+    },
+    {
+      "key": "gemini-3-7-flash",
+      "name": "Gemini 3.7 Flash",
+      "agree_whole": 89.0,
+      "agree_crit": 93.3,
+      "whole": {
+        "fp": 13,
+        "fn": 1,
+        "fails": 96,
+        "passes": 66
+      },
+      "crit": {
+        "fp": 46,
+        "fn": 6,
+        "fails": 199,
+        "passes": 886
+      }
+    },
+    {
+      "key": "kimi-k3",
+      "name": "Kimi K3",
+      "agree_whole": 84.9,
+      "agree_crit": 92.4,
+      "whole": {
+        "fp": 21,
+        "fn": 0,
+        "fails": 96,
+        "passes": 66
+      },
+      "crit": {
+        "fp": 54,
+        "fn": 8,
+        "fails": 199,
+        "passes": 886
+      }
+    }
+  ],
+  "human_agreement": {
+    "agree_whole": 94.2,
+    "agree_crit": 95.8
+  },
+  "judge_study": {
+    "tasks": 43,
+    "responses": 172
+  },
+  "task_and_rubric_breakdowns": {
+    "task_count": 100,
+    "input_file_count": 435,
+    "file_types": [
+      {
+        "kind": "Image files",
+        "tasks": 25,
+        "files": 55
+      },
+      {
+        "kind": "PDF",
+        "tasks": 66,
+        "files": 138
+      },
+      {
+        "kind": "Audio",
+        "tasks": 9,
+        "files": 9
+      },
+      {
+        "kind": "Video",
+        "tasks": 1,
+        "files": 1
+      },
+      {
+        "kind": "Tables / spreadsheets",
+        "tasks": 60,
+        "files": 88
+      },
+      {
+        "kind": "Text / structured text",
+        "tasks": 45,
+        "files": 144
+      }
+    ],
+    "image_only_pdf_files": 8,
+    "image_only_pdf_tasks": 4,
+    "image_or_image_only_pdf_tasks": 28,
+    "image_audio_video_union_tasks": 33,
+    "nontext_evidence_tasks": 30,
+    "nontext_evidence_rubric_items": 90,
+    "primary_domains": {
+      "engineering": 46,
+      "medicine": 13,
+      "science": 7,
+      "aviation": 1,
+      "law": 7,
+      "public_records": 14,
+      "finance": 9,
+      "other": 3
+    },
+    "rubric_categories": [
+      {
+        "category": "Fact / rule identification",
+        "items": 217,
+        "item_share": 0.3273001508295626,
+        "passed": 6495,
+        "judgements": 7161,
+        "pass_rate": 0.9069962295768748,
+        "models": [
+          {
+            "model": "anthropic/claude-opus-5",
+            "passed": 615,
+            "judgements": 651,
+            "pass_rate": 0.9447004608294931
+          },
+          {
+            "model": "anthropic/claude-opus-5.5",
+            "passed": 645,
+            "judgements": 651,
+            "pass_rate": 0.9907834101382489
+          },
+          {
+            "model": "anthropic/claude-sonnet-5",
+            "passed": 567,
+            "judgements": 651,
+            "pass_rate": 0.8709677419354839
+          },
+          {
+            "model": "google/gemini-3.8-flash",
+            "passed": 596,
+            "judgements": 651,
+            "pass_rate": 0.9155145929339478
+          },
+          {
+            "model": "google/gemini-4-argon",
+            "passed": 636,
+            "judgements": 651,
+            "pass_rate": 0.9769585253456221
+          },
+          {
+            "model": "gpt-5.6-sol",
+            "passed": 627,
+            "judgements": 651,
+            "pass_rate": 0.9631336405529954
+          },
+          {
+            "model": "gpt-6-astra",
+            "passed": 637,
+            "judgements": 651,
+            "pass_rate": 0.978494623655914
+          },
+          {
+            "model": "gpt-6.1-sol",
+            "passed": 632,
+            "judgements": 651,
+            "pass_rate": 0.9708141321044547
+          },
+          {
+            "model": "moonshotai/kimi-k3",
+            "passed": 579,
+            "judgements": 651,
+            "pass_rate": 0.8894009216589862
+          },
+          {
+            "model": "thinkingmachines/inkling",
+            "passed": 355,
+            "judgements": 651,
+            "pass_rate": 0.5453149001536098
+          },
+          {
+            "model": "x-ai/grok-4.6",
+            "passed": 606,
+            "judgements": 651,
+            "pass_rate": 0.9308755760368663
+          }
+        ]
+      },
+      {
+        "category": "Joined application / comparison",
+        "items": 289,
+        "item_share": 0.4358974358974359,
+        "passed": 8107,
+        "judgements": 9537,
+        "pass_rate": 0.8500576701268743,
+        "models": [
+          {
+            "model": "anthropic/claude-opus-5",
+            "passed": 798,
+            "judgements": 867,
+            "pass_rate": 0.9204152249134948
+          },
+          {
+            "model": "anthropic/claude-opus-5.5",
+            "passed": 845,
+            "judgements": 867,
+            "pass_rate": 0.9746251441753172
+          },
+          {
+            "model": "anthropic/claude-sonnet-5",
+            "passed": 629,
+            "judgements": 867,
+            "pass_rate": 0.7254901960784313
+          },
+          {
+            "model": "google/gemini-3.8-flash",
+            "passed": 766,
+            "judgements": 867,
+            "pass_rate": 0.8835063437139562
+          },
+          {
+            "model": "google/gemini-4-argon",
+            "passed": 848,
+            "judgements": 867,
+            "pass_rate": 0.9780853517877739
+          },
+          {
+            "model": "gpt-5.6-sol",
+            "passed": 782,
+            "judgements": 867,
+            "pass_rate": 0.9019607843137255
+          },
+          {
+            "model": "gpt-6-astra",
+            "passed": 836,
+            "judgements": 867,
+            "pass_rate": 0.9642445213379469
+          },
+          {
+            "model": "gpt-6.1-sol",
+            "passed": 819,
+            "judgements": 867,
+            "pass_rate": 0.9446366782006921
+          },
+          {
+            "model": "moonshotai/kimi-k3",
+            "passed": 679,
+            "judgements": 867,
+            "pass_rate": 0.7831603229527105
+          },
+          {
+            "model": "thinkingmachines/inkling",
+            "passed": 313,
+            "judgements": 867,
+            "pass_rate": 0.36101499423298733
+          },
+          {
+            "model": "x-ai/grok-4.6",
+            "passed": 792,
+            "judgements": 867,
+            "pass_rate": 0.9134948096885813
+          }
+        ]
+      },
+      {
+        "category": "Calculation / transformation",
+        "items": 86,
+        "item_share": 0.1297134238310709,
+        "passed": 2445,
+        "judgements": 2838,
+        "pass_rate": 0.861522198731501,
+        "models": [
+          {
+            "model": "anthropic/claude-opus-5",
+            "passed": 238,
+            "judgements": 258,
+            "pass_rate": 0.9224806201550387
+          },
+          {
+            "model": "anthropic/claude-opus-5.5",
+            "passed": 251,
+            "judgements": 258,
+            "pass_rate": 0.9728682170542635
+          },
+          {
+            "model": "anthropic/claude-sonnet-5",
+            "passed": 197,
+            "judgements": 258,
+            "pass_rate": 0.7635658914728682
+          },
+          {
+            "model": "google/gemini-3.8-flash",
+            "passed": 237,
+            "judgements": 258,
+            "pass_rate": 0.9186046511627907
+          },
+          {
+            "model": "google/gemini-4-argon",
+            "passed": 252,
+            "judgements": 258,
+            "pass_rate": 0.9767441860465116
+          },
+          {
+            "model": "gpt-5.6-sol",
+            "passed": 224,
+            "judgements": 258,
+            "pass_rate": 0.8682170542635659
+          },
+          {
+            "model": "gpt-6-astra",
+            "passed": 245,
+            "judgements": 258,
+            "pass_rate": 0.9496124031007752
+          },
+          {
+            "model": "gpt-6.1-sol",
+            "passed": 245,
+            "judgements": 258,
+            "pass_rate": 0.9496124031007752
+          },
+          {
+            "model": "moonshotai/kimi-k3",
+            "passed": 207,
+            "judgements": 258,
+            "pass_rate": 0.8023255813953488
+          },
+          {
+            "model": "thinkingmachines/inkling",
+            "passed": 108,
+            "judgements": 258,
+            "pass_rate": 0.4186046511627907
+          },
+          {
+            "model": "x-ai/grok-4.6",
+            "passed": 241,
+            "judgements": 258,
+            "pass_rate": 0.9341085271317829
+          }
+        ]
+      },
+      {
+        "category": "Grounding / evidence limits",
+        "items": 71,
+        "item_share": 0.10708898944193061,
+        "passed": 2004,
+        "judgements": 2343,
+        "pass_rate": 0.8553137003841229,
+        "models": [
+          {
+            "model": "anthropic/claude-opus-5",
+            "passed": 186,
+            "judgements": 213,
+            "pass_rate": 0.8732394366197183
+          },
+          {
+            "model": "anthropic/claude-opus-5.5",
+            "passed": 196,
+            "judgements": 213,
+            "pass_rate": 0.92018779342723
+          },
+          {
+            "model": "anthropic/claude-sonnet-5",
+            "passed": 166,
+            "judgements": 213,
+            "pass_rate": 0.7793427230046949
+          },
+          {
+            "model": "google/gemini-3.8-flash",
+            "passed": 167,
+            "judgements": 213,
+            "pass_rate": 0.784037558685446
+          },
+          {
+            "model": "google/gemini-4-argon",
+            "passed": 200,
+            "judgements": 213,
+            "pass_rate": 0.9389671361502347
+          },
+          {
+            "model": "gpt-5.6-sol",
+            "passed": 199,
+            "judgements": 213,
+            "pass_rate": 0.9342723004694836
+          },
+          {
+            "model": "gpt-6-astra",
+            "passed": 202,
+            "judgements": 213,
+            "pass_rate": 0.9483568075117371
+          },
+          {
+            "model": "gpt-6.1-sol",
+            "passed": 200,
+            "judgements": 213,
+            "pass_rate": 0.9389671361502347
+          },
+          {
+            "model": "moonshotai/kimi-k3",
+            "passed": 171,
+            "judgements": 213,
+            "pass_rate": 0.8028169014084507
+          },
+          {
+            "model": "thinkingmachines/inkling",
+            "passed": 129,
+            "judgements": 213,
+            "pass_rate": 0.6056338028169014
+          },
+          {
+            "model": "x-ai/grok-4.6",
+            "passed": 188,
+            "judgements": 213,
+            "pass_rate": 0.8826291079812206
+          }
+        ]
+      }
+    ],
+    "category_definitions": {
+      "Fact / rule identification": "Remaining criteria with only one source environment, including direct observations and rule or fact identification.",
+      "Joined application / comparison": "Remaining criteria supported by both workspace and web sources, including comparisons and application of public rules to the local case.",
+      "Calculation / transformation": "Required derived numerical values, arithmetic or numerical methods, time or unit conversions, or numerical feasibility checks.",
+      "Grounding / evidence limits": "Explicit source/locator requirements or requirements to limit claims to what the available evidence establishes."
+    },
+    "category_precedence": [
+      "Calculation / transformation",
+      "Grounding / evidence limits",
+      "Joined application / comparison",
+      "Fact / rule identification"
+    ],
+    "provenance": "Primary domains are original author-provided submission labels. Rubric criterionType is empty for all 663 items. Four reporting categories are exploratory post-hoc coding of primary requirement and original source-support graph, not original labels, a new grading rubric, or a causal failure taxonomy. Calculation and grounding/limits labels were reviewed against criterion text. Remainder splits on whether the linked source graph uses both workspace and web sources.",
+    "rate_definition": "Pooled pass rate of saved per-item boolean judgments across all 11 native-search models, all 100 tasks and all 3 trials per model/task. Each of 663 items has 33 judgments. Categories have different difficulty and composition; rates are descriptive and do not establish causes.",
+    "scoring_basis": "Saved per-item outcomes for non-perfect attempts; validated full-credit records imply every unit-weight item passed. No new grading.",
+    "inventory_definition": "Delivered environment/artifacts files excluding the standard answer_format.md. Original submission URLs without delivered local files are excluded. Image-only PDF means no extractable text and at least one embedded raster image; PDFs with an OCR layer are not identified as image-only."
+  },
+  "cost_note": "Eight model costs are recorded. Gemini 4 Argon, Gemini 3.8 Flash and Kimi K3 costs use saved input, cached-input and output token counts at the November 3, 2026 rate snapshot. Agent cost excludes failed reruns, grading, tools and infrastructure.",
+  "cohort_note": "The 1,269 earlier capped search runs are retained separately and are not pooled with the 1,350 uncapped runs.",
+  "tasks": 100,
+  "runs": 5919,
+  "native_runs": 3300,
+  "ablation_runs": 2619,
+  "uncapped_ablation_runs": 1350,
+  "legacy_capped_ablation_runs": 1269,
+  "current_evaluation_runs": 4650,
+  "metrics": [
+    {
+      "model": "anthropic/claude-opus-5",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.9224380066666666,
+      "full_credit_pass_at_1": 0.6966666666666667,
+      "pass_power_k": {
+        "1": 0.6966666666666667,
+        "2": 0.59,
+        "3": 0.51
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 17,
+        "1": 8,
+        "2": 24,
+        "3": 51
+      }
+    },
+    {
+      "model": "anthropic/claude-opus-5.5",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.96980951,
+      "full_credit_pass_at_1": 0.86,
+      "pass_power_k": {
+        "1": 0.86,
+        "2": 0.8,
+        "3": 0.76
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 6,
+        "1": 6,
+        "2": 12,
+        "3": 76
+      }
+    },
+    {
+      "model": "anthropic/claude-sonnet-5",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.7452871033333334,
+      "full_credit_pass_at_1": 0.3433333333333333,
+      "pass_power_k": {
+        "1": 0.3433333333333333,
+        "2": 0.17666666666666667,
+        "3": 0.11
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 39,
+        "1": 30,
+        "2": 20,
+        "3": 11
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "exa",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8852179432624113,
+      "full_credit_pass_at_1": 0.5460992907801419,
+      "pass_power_k": {
+        "1": 0.5460992907801419,
+        "2": 0.475177304964539,
+        "3": 0.425531914893617
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 17,
+        "1": 3,
+        "2": 7,
+        "3": 20
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "exa",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.89926358,
+      "full_credit_pass_at_1": 0.5866666666666667,
+      "pass_power_k": {
+        "1": 0.5866666666666667,
+        "2": 0.5,
+        "3": 0.46
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 14,
+        "1": 7,
+        "2": 6,
+        "3": 23
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "firecrawl",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8536526382978723,
+      "full_credit_pass_at_1": 0.44680851063829785,
+      "pass_power_k": {
+        "1": 0.44680851063829785,
+        "2": 0.3262411347517731,
+        "3": 0.2765957446808511
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 17,
+        "1": 10,
+        "2": 7,
+        "3": 13
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "firecrawl",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.9175541,
+      "full_credit_pass_at_1": 0.6266666666666667,
+      "pass_power_k": {
+        "1": 0.6266666666666667,
+        "2": 0.5266666666666666,
+        "3": 0.48
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 11,
+        "1": 8,
+        "2": 7,
+        "3": 24
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.8724971966666667,
+      "full_credit_pass_at_1": 0.58,
+      "pass_power_k": {
+        "1": 0.58,
+        "2": 0.46,
+        "3": 0.38
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 26,
+        "1": 12,
+        "2": 24,
+        "3": 38
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "parallel",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8851616595744681,
+      "full_credit_pass_at_1": 0.5673758865248227,
+      "pass_power_k": {
+        "1": 0.5673758865248227,
+        "2": 0.475177304964539,
+        "3": 0.44680851063829785
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 13,
+        "1": 9,
+        "2": 4,
+        "3": 21
+      }
+    },
+    {
+      "model": "google/gemini-3.8-flash",
+      "condition": "parallel",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.9194328933333333,
+      "full_credit_pass_at_1": 0.6466666666666666,
+      "pass_power_k": {
+        "1": 0.6466666666666666,
+        "2": 0.56,
+        "3": 0.5
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 12,
+        "1": 4,
+        "2": 9,
+        "3": 25
+      }
+    },
+    {
+      "model": "google/gemini-4-argon",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.975123,
+      "full_credit_pass_at_1": 0.8666666666666667,
+      "pass_power_k": {
+        "1": 0.8666666666666667,
+        "2": 0.8266666666666667,
+        "3": 0.81
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 7,
+        "1": 7,
+        "2": 5,
+        "3": 81
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "exa",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.9034142765957447,
+      "full_credit_pass_at_1": 0.6382978723404256,
+      "pass_power_k": {
+        "1": 0.6382978723404256,
+        "2": 0.5035460992907801,
+        "3": 0.40425531914893614
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 9,
+        "1": 5,
+        "2": 14,
+        "3": 19
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "exa",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.9261211533333333,
+      "full_credit_pass_at_1": 0.7266666666666667,
+      "pass_power_k": {
+        "1": 0.7266666666666667,
+        "2": 0.6466666666666666,
+        "3": 0.6
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 8,
+        "1": 5,
+        "2": 7,
+        "3": 30
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "firecrawl",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8824806028368795,
+      "full_credit_pass_at_1": 0.5957446808510638,
+      "pass_power_k": {
+        "1": 0.5957446808510638,
+        "2": 0.48936170212765956,
+        "3": 0.425531914893617
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 12,
+        "1": 6,
+        "2": 9,
+        "3": 20
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "firecrawl",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.91654498,
+      "full_credit_pass_at_1": 0.7,
+      "pass_power_k": {
+        "1": 0.7,
+        "2": 0.5933333333333334,
+        "3": 0.54
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 7,
+        "1": 8,
+        "2": 8,
+        "3": 27
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.9074536733333334,
+      "full_credit_pass_at_1": 0.68,
+      "pass_power_k": {
+        "1": 0.68,
+        "2": 0.5533333333333333,
+        "3": 0.47
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 15,
+        "1": 13,
+        "2": 25,
+        "3": 47
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "parallel",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8948098156028369,
+      "full_credit_pass_at_1": 0.6099290780141844,
+      "pass_power_k": {
+        "1": 0.6099290780141844,
+        "2": 0.5106382978723404,
+        "3": 0.44680851063829785
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 12,
+        "1": 5,
+        "2": 9,
+        "3": 21
+      }
+    },
+    {
+      "model": "gpt-5.6-sol",
+      "condition": "parallel",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.9187063466666666,
+      "full_credit_pass_at_1": 0.7066666666666667,
+      "pass_power_k": {
+        "1": 0.7066666666666667,
+        "2": 0.62,
+        "3": 0.58
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 8,
+        "1": 7,
+        "2": 6,
+        "3": 29
+      }
+    },
+    {
+      "model": "gpt-6-astra",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.9581732766666666,
+      "full_credit_pass_at_1": 0.82,
+      "pass_power_k": {
+        "1": 0.82,
+        "2": 0.7833333333333333,
+        "3": 0.76
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 13,
+        "1": 4,
+        "2": 7,
+        "3": 76
+      }
+    },
+    {
+      "model": "gpt-6.1-sol",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.94528174,
+      "full_credit_pass_at_1": 0.7933333333333333,
+      "pass_power_k": {
+        "1": 0.7933333333333333,
+        "2": 0.75,
+        "3": 0.73
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 14,
+        "1": 7,
+        "2": 6,
+        "3": 73
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "exa",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.7981263191489362,
+      "full_credit_pass_at_1": 0.375886524822695,
+      "pass_power_k": {
+        "1": 0.375886524822695,
+        "2": 0.23404255319148937,
+        "3": 0.19148936170212766
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 18,
+        "1": 14,
+        "2": 6,
+        "3": 9
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "exa",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.8570790266666667,
+      "full_credit_pass_at_1": 0.5333333333333333,
+      "pass_power_k": {
+        "1": 0.5333333333333333,
+        "2": 0.41333333333333333,
+        "3": 0.34
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 15,
+        "1": 7,
+        "2": 11,
+        "3": 17
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "firecrawl",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.7683709078014185,
+      "full_credit_pass_at_1": 0.36879432624113473,
+      "pass_power_k": {
+        "1": 0.36879432624113473,
+        "2": 0.2553191489361702,
+        "3": 0.19148936170212766
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 22,
+        "1": 7,
+        "2": 9,
+        "3": 9
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "firecrawl",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.84177482,
+      "full_credit_pass_at_1": 0.46,
+      "pass_power_k": {
+        "1": 0.46,
+        "2": 0.36,
+        "3": 0.3
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 20,
+        "1": 6,
+        "2": 9,
+        "3": 15
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.8092198866666667,
+      "full_credit_pass_at_1": 0.41,
+      "pass_power_k": {
+        "1": 0.41,
+        "2": 0.2733333333333333,
+        "3": 0.23
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 36,
+        "1": 28,
+        "2": 13,
+        "3": 23
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "parallel",
+      "cohort": "legacy_capped",
+      "tasks": 47,
+      "runs": 141,
+      "mean_reward": 0.8110883475177305,
+      "full_credit_pass_at_1": 0.375886524822695,
+      "pass_power_k": {
+        "1": 0.375886524822695,
+        "2": 0.2553191489361702,
+        "3": 0.19148936170212766
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 21,
+        "1": 8,
+        "2": 9,
+        "3": 9
+      }
+    },
+    {
+      "model": "moonshotai/kimi-k3",
+      "condition": "parallel",
+      "cohort": "uncapped_20261104",
+      "tasks": 50,
+      "runs": 150,
+      "mean_reward": 0.85915192,
+      "full_credit_pass_at_1": 0.5266666666666666,
+      "pass_power_k": {
+        "1": 0.5266666666666666,
+        "2": 0.3933333333333333,
+        "3": 0.34
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 13,
+        "1": 12,
+        "2": 8,
+        "3": 17
+      }
+    },
+    {
+      "model": "thinkingmachines/inkling",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.41612396,
+      "full_credit_pass_at_1": 0.03,
+      "pass_power_k": {
+        "1": 0.03,
+        "2": 0.0,
+        "3": 0.0
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 91,
+        "1": 9
+      }
+    },
+    {
+      "model": "x-ai/grok-4.6",
+      "condition": "native",
+      "cohort": "native",
+      "tasks": 100,
+      "runs": 300,
+      "mean_reward": 0.91473413,
+      "full_credit_pass_at_1": 0.69,
+      "pass_power_k": {
+        "1": 0.69,
+        "2": 0.5933333333333334,
+        "3": 0.54
+      },
+      "perfect_attempts_per_task_histogram": {
+        "0": 17,
+        "1": 13,
+        "2": 16,
+        "3": 54
+      }
+    }
+  ],
+  "no_new_grading_or_model_calls": true,
+  "method": "Metrics are aggregated from retained completed run indexes. Cohort labels distinguish native, current uncapped and historical capped runs; capped and uncapped results are never pooled."
+}

run-index-20261104.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a01396b0888140a8206e6405703f22430dda88be1eac8857e7423bdfec1a372d
+size 195686

run-index-legacy-4269.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53da304df8a44aa33b53b21460576910a25a2f9f476fb4dcb81ec6390bf0d80e
+size 505283

run-index.jsonl.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:53da304df8a44aa33b53b21460576910a25a2f9f476fb4dcb81ec6390bf0d80e
-size 505283
+oid sha256:bfe956833752bbb4e02fa93612dc840eecde68a9f3bd3949d29b60424095e5af
+size 700765