{
 "chart": {
  "categories": [
   "Thinking",
   "No-think",
   "Next-op",
   "Surface",
   "Random"
  ],
  "live_child_recall_at_4_percent": [
   25.067,
   30.315,
   22.8398,
   27.7803,
   26.2748
  ],
  "macro_within_task_auroc_percent": [
   50.6443,
   55.5627,
   50.365,
   51.9021,
   50.5841
  ]
 },
 "data_integrity": {
  "common_behavior_signatures": 120,
  "cross_split_collisions": 0,
  "exact_receipts_recomputed": 120,
  "passed": true,
  "task_oracle_pairs_checked": 120
 },
 "experiment_id": "qwen35_4b_partial_structure_search",
 "full_brute_primary": {
  "aggregate_accounting": {
   "behavior_states_inserted": 29163087,
   "case_operation_applications": 254364663,
   "full_skeletons_yielded": 0,
   "invalid_vector_transitions": 129814,
   "parameter_fills_yielded": 0,
   "prefix_rows_emitted": 0,
   "pruning_case_candidate_pipelines": 5018189,
   "successful_concrete_pipelines": 1716,
   "successful_type_skeletons": 1074,
   "transition_requests": 227198208,
   "validation_case_operation_applications": 27166455,
   "validation_pipelines_checked": 5018189,
   "vector_transitions_computed": 227198208
  },
  "logical_type_skeleton_leaves": 62914560,
  "parallel_wall_seconds": 111.84706237912178,
  "path_coverage_rate": 1.0,
  "selected_hidden_success_rate": 0.9333333333333333,
  "task_count": 60
 },
 "model_resources": {
  "scoring_summaries": {
   "nextop": {
    "accounting": {
     "completions": 450,
     "discarded_sampled_tokens": 0,
     "injected_prompt_tokens": 1350,
     "prefill_tokens": 286871,
     "requests": 450,
     "retained_thinking_tokens": 0,
     "sampled_tokens": 450,
     "total_model_tokens": 287321
    },
    "label_token_ids": {
     "A": 32,
     "B": 33,
     "C": 34,
     "D": 35,
     "E": 36,
     "F": 37,
     "G": 38,
     "H": 39,
     "I": 40,
     "J": 41,
     "K": 42,
     "L": 43,
     "M": 44,
     "N": 45,
     "O": 46,
     "P": 47
    },
    "method": "next_operation_likelihood",
    "run_seed": 1703,
    "vllm_batch_calls": 1
   },
   "nothink": {
    "accounting": {
     "completions": 7200,
     "discarded_sampled_tokens": 0,
     "injected_prompt_tokens": 21600,
     "prefill_tokens": 4111664,
     "requests": 7200,
     "retained_thinking_tokens": 0,
     "sampled_tokens": 7200,
     "total_model_tokens": 4118864
    },
    "label_semantics": {
     "A": "viable",
     "B": "not_viable"
    },
    "label_token_ids": {
     "A": 32,
     "B": 33
    },
    "method": "no_think_p_viable",
    "run_seed": 1702,
    "vllm_batch_calls": 1
   },
   "task_shuffled_thinking": {
    "accounting": {
     "completions": 2400,
     "discarded_sampled_tokens": 0,
     "injected_prompt_tokens": 6000,
     "prefill_tokens": 1664976,
     "requests": 2400,
     "retained_thinking_tokens": 307200,
     "sampled_tokens": 308400,
     "total_model_tokens": 1973376
    },
    "label_semantics": {
     "A": "viable",
     "B": "not_viable"
    },
    "label_token_ids": {
     "A": 32,
     "B": 33
    },
    "method": "thinking_p_viable",
    "run_seed": 1704,
    "thinking_answer_suffix_token_ids": [
     248069,
     271,
     15666,
     25,
     220
    ],
    "thinking_budget": 256,
    "vllm_batch_calls": 2
   },
   "thinking": {
    "accounting": {
     "completions": 14400,
     "discarded_sampled_tokens": 0,
     "injected_prompt_tokens": 36000,
     "prefill_tokens": 10030528,
     "requests": 14400,
     "retained_thinking_tokens": 1843200,
     "sampled_tokens": 1850400,
     "total_model_tokens": 11880928
    },
    "label_semantics": {
     "A": "viable",
     "B": "not_viable"
    },
    "label_token_ids": {
     "A": 32,
     "B": 33
    },
    "method": "thinking_p_viable",
    "run_seed": 1701,
    "thinking_answer_suffix_token_ids": [
     248069,
     271,
     15666,
     25,
     220
    ],
    "thinking_budget": 256,
    "vllm_batch_calls": 2
   }
  },
  "wall_seconds": 2563.240444508381
 },
 "oracle_development_gate": {
  "live_density_by_length": {
   "1": 0.11979166666666667,
   "2": 0.013997395833333334,
   "3": 0.0013224283854166667,
   "4": 0.000118255615234375,
   "5": 7.788340250651041e-06
  },
  "mean_compression": 262144.0,
  "mean_oracle_wall_seconds": 13.776590572282052,
  "n": 12,
  "path_rate": 1.0,
  "selected_rate": 1.0
 },
 "raw_trace_compaction": {
  "artifacts": [
   {
    "compact_path": "runs/calibration_thinking.jsonl",
    "compact_sha256": "15a2c30b5c20a072e54461e5e16ad8ce771fa3e377cd11b6ce76e11894566541",
    "compact_size_bytes": 2249635,
    "external_raw_path": "large_artifacts/qwen35_4b_partial_structure_search/calibration/calibration_thinking.raw.jsonl",
    "raw_sha256": "c18f1505318def01d1b12e96eba57d551063ff747ccc5d88bf796ba77c2b0914",
    "raw_size_bytes": 110639758,
    "row_count": 7200,
    "tag": "thinking"
   },
   {
    "compact_path": "runs/calibration_nothink.jsonl",
    "compact_sha256": "468cd973d62e53b5bd7c87583992f293cd59012656540781bd7131f419bf4f30",
    "compact_size_bytes": 2207338,
    "external_raw_path": "large_artifacts/qwen35_4b_partial_structure_search/calibration/calibration_nothink.raw.jsonl",
    "raw_sha256": "7172e7cce43a48153247d965c8f2f2b24c6e46608a6371c031af7ef272e155d0",
    "raw_size_bytes": 43940086,
    "row_count": 7200,
    "tag": "nothink"
   },
   {
    "compact_path": "runs/calibration_nextop.jsonl",
    "compact_sha256": "20d2d75e7a6c7bb90e1977723392067d3b89230f8502bb26b8dc234e712d7c5f",
    "compact_size_bytes": 364386,
    "external_raw_path": "large_artifacts/qwen35_4b_partial_structure_search/calibration/calibration_nextop.raw.jsonl",
    "raw_sha256": "ca8a28cdbaf72d21c570de8549f1e1814b8cf77d3e0002684876eea45c81713a",
    "raw_size_bytes": 4061100,
    "row_count": 450,
    "tag": "nextop"
   },
   {
    "compact_path": "runs/calibration_task_shuffled.jsonl",
    "compact_sha256": "bfa1309958790a58fb46bb6bc508b5da70b96cb6b01f19df616f4927e7952330",
    "compact_size_bytes": 509371,
    "external_raw_path": "large_artifacts/qwen35_4b_partial_structure_search/calibration/calibration_task_shuffled.raw.jsonl",
    "raw_sha256": "1be073a03242135ea72c5e16a16033a0321ba2cc2a8c5fdcd021ed77993bbc1e",
    "raw_size_bytes": 18535407,
    "row_count": 1200,
    "tag": "task_shuffled"
   }
  ],
  "operation": "external_raw_archive_with_analysis_complete_compact_replacement",
  "receipt_before_compaction_sha256": "da9042fd28c83d82793ff014a5d9e09276ae9ccd7c500aee852376e683a3f89b",
  "schema_version": 1
 },
 "recognition_calibration": {
  "gate_checks": {
   "auroc_lift_lcb": false,
   "auroc_lift_point": false,
   "macro_auroc_lcb": false,
   "macro_auroc_point": false,
   "recall_lift_lcb": false,
   "recall_lift_point": false
  },
  "methods": {
   "nextop": {
    "best_live_rank": {
     "ci_high": 7.707185019841268,
     "ci_low": 6.234718501984126,
     "estimate": 6.983184523809523,
     "n_tasks": 48,
     "reps": 10000
    },
    "live_recall_at_beam": {
     "ci_high": 0.2872290426587301,
     "ci_low": 0.17358444940476192,
     "estimate": 0.22839781746031748,
     "n_tasks": 48,
     "reps": 10000
    },
    "macro_auroc": {
     "ci_high": 0.5437923561449874,
     "ci_low": 0.46474390754672035,
     "estimate": 0.5036495776354728,
     "n_tasks": 48,
     "reps": 10000
    },
    "pooled_auroc_diagnostic": 0.5214402467884758
   },
   "nothink": {
    "best_live_rank": {
     "ci_high": 7.18322048611111,
     "ci_low": 5.8816443452380955,
     "estimate": 6.527232142857144,
     "n_tasks": 48,
     "reps": 10000
    },
    "live_recall_at_beam": {
     "ci_high": 0.3554575892857142,
     "ci_low": 0.251437251984127,
     "estimate": 0.3031498015873016,
     "n_tasks": 48,
     "reps": 10000
    },
    "macro_auroc": {
     "ci_high": 0.5941857582314344,
     "ci_low": 0.5159039722849043,
     "estimate": 0.5556273839439984,
     "n_tasks": 48,
     "reps": 10000
    },
    "pooled_auroc_diagnostic": 0.5802208821592498
   },
   "random": {
    "best_live_rank": {
     "ci_high": 7.45883060515873,
     "ci_low": 6.445979662698413,
     "estimate": 6.952579365079365,
     "n_tasks": 48,
     "reps": 10000
    },
    "live_recall_at_beam": {
     "ci_high": 0.30823660714285717,
     "ci_low": 0.21914558531746034,
     "estimate": 0.2627480158730159,
     "n_tasks": 48,
     "reps": 10000
    },
    "macro_auroc": {
     "ci_high": 0.5356814611179622,
     "ci_low": 0.47765112207166277,
     "estimate": 0.5058407500552705,
     "n_tasks": 48,
     "reps": 10000
    },
    "pooled_auroc_diagnostic": 0.4977464243432232
   },
   "surface": {
    "best_live_rank": {
     "ci_high": 8.120792410714285,
     "ci_low": 6.618921130952382,
     "estimate": 7.370039682539683,
     "n_tasks": 48,
     "reps": 10000
    },
    "live_recall_at_beam": {
     "ci_high": 0.3395343501984127,
     "ci_low": 0.21621961805555553,
     "estimate": 0.27780257936507935,
     "n_tasks": 48,
     "reps": 10000
    },
    "macro_auroc": {
     "ci_high": 0.5534554295052464,
     "ci_low": 0.48498958799685854,
     "estimate": 0.5190213903708779,
     "n_tasks": 48,
     "reps": 10000
    },
    "pooled_auroc_diagnostic": 0.5361185217091446
   },
   "thinking": {
    "best_live_rank": {
     "ci_high": 7.689574652777776,
     "ci_low": 6.521121031746032,
     "estimate": 7.101934523809525,
     "n_tasks": 48,
     "reps": 10000
    },
    "live_recall_at_beam": {
     "ci_high": 0.2987109375,
     "ci_low": 0.20223214285714286,
     "estimate": 0.2506696428571428,
     "n_tasks": 48,
     "reps": 10000
    },
    "macro_auroc": {
     "ci_high": 0.5434581428192307,
     "ci_low": 0.47036138443796427,
     "estimate": 0.5064432265726517,
     "n_tasks": 48,
     "reps": 10000
    },
    "pooled_auroc_diagnostic": 0.5573586512282781
   }
  },
  "n_candidates": 7200,
  "n_tasks": 48,
  "strongest_auc_baseline": "nothink",
  "strongest_recall_baseline": "nothink",
  "task_shuffle_canary": {
   "n_candidates": 1200,
   "n_tasks": 8,
   "original_macro_auroc": {
    "ci_high": 0.5512256532067837,
    "ci_low": 0.3652553643415549,
    "estimate": 0.45016614326130905,
    "n_tasks": 8,
    "reps": 10000
   },
   "original_minus_shuffled": {
    "ci_high": 0.08986450248757948,
    "ci_low": -0.14582833746772156,
    "estimate": -0.025360848875355224,
    "n_tasks": 8,
    "reps": 10000
   },
   "task_shuffled_macro_auroc": {
    "ci_high": 0.5585101871191142,
    "ci_low": 0.40512706248526964,
    "estimate": 0.4755269921366643,
    "n_tasks": 8,
    "reps": 10000
   }
  },
  "thinking_forced_close_rate": 1.0,
  "thinking_vs_strongest_auc": {
   "ci_high": -0.009503187082277601,
   "ci_low": -0.08973016391992075,
   "estimate": -0.04918415737134662,
   "n_tasks": 48,
   "reps": 10000
  },
  "thinking_vs_strongest_recall": {
   "ci_high": 0.007142857142857141,
   "ci_low": -0.11046874999999999,
   "estimate": -0.05248015873015873,
   "n_tasks": 48,
   "reps": 10000
  }
 },
 "schema_version": 1,
 "verdict": {
  "banking_run": false,
  "classification": "G1_unreadable_partial_state",
  "primary_search_run": false,
  "reason": "thinking viability failed every preregistered recognition/actionability check",
  "recognition_gate_passed": false
 }
}