{
  "assigned": 2880,
  "observed": 2876,
  "missing": 4,
  "actual_started_utc": "2026-10-09T21:03:17.631184+00:00",
  "actual_ended_utc": "2026-10-09T23:20:05.861241+00:00",
  "conditions": [
    {
      "model": "qwen2.5-coder:0.5b",
      "condition": "single-open",
      "assigned": 360,
      "observed": 360,
      "correct": 62,
      "success_fraction_observed": 0.17222222222222222,
      "assigned_correctness_bounds": [
        0.17222222222222222,
        0.17222222222222222
      ],
      "wrong_solved_claims": 295,
      "eligible_solved_claims": 357,
      "wrong_claim_fraction": 0.8263305322128851,
      "wrong_controller_acceptance": 295,
      "actual_final_checks": 230,
      "mean_calls": 1.025,
      "mean_tokens": 495.2611111111111,
      "mean_wall_seconds": 0.5743388888880469,
      "mean_api_seconds": 0.567680555554883
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "condition": "single-contract",
      "assigned": 360,
      "observed": 360,
      "correct": 59,
      "success_fraction_observed": 0.1638888888888889,
      "assigned_correctness_bounds": [
        0.1638888888888889,
        0.1638888888888889
      ],
      "wrong_solved_claims": 684,
      "eligible_solved_claims": 743,
      "wrong_claim_fraction": 0.9205921938088829,
      "wrong_controller_acceptance": 164,
      "actual_final_checks": 229,
      "mean_calls": 2.1333333333333333,
      "mean_tokens": 1787.0555555555557,
      "mean_wall_seconds": 0.9017027777784582,
      "mean_api_seconds": 0.8834361111111422
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "condition": "staged-open",
      "assigned": 360,
      "observed": 360,
      "correct": 62,
      "success_fraction_observed": 0.17222222222222222,
      "assigned_correctness_bounds": [
        0.17222222222222222,
        0.17222222222222222
      ],
      "wrong_solved_claims": 132,
      "eligible_solved_claims": 135,
      "wrong_claim_fraction": 0.9777777777777777,
      "wrong_controller_acceptance": 132,
      "actual_final_checks": 240,
      "mean_calls": 4,
      "mean_tokens": 2995.6222222222223,
      "mean_wall_seconds": 1.7824138888891967,
      "mean_api_seconds": 1.7505388888890998
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "condition": "staged-contract",
      "assigned": 360,
      "observed": 359,
      "correct": 63,
      "success_fraction_observed": 0.17548746518105848,
      "assigned_correctness_bounds": [
        0.175,
        0.17777777777777778
      ],
      "wrong_solved_claims": 162,
      "eligible_solved_claims": 173,
      "wrong_claim_fraction": 0.9364161849710982,
      "wrong_controller_acceptance": 95,
      "actual_final_checks": 243,
      "mean_calls": 4,
      "mean_tokens": 3014.6295264623955,
      "mean_wall_seconds": 1.8203649025076085,
      "mean_api_seconds": 1.7883175487469876
    },
    {
      "model": "qwen2.5:0.5b",
      "condition": "single-open",
      "assigned": 360,
      "observed": 360,
      "correct": 4,
      "success_fraction_observed": 0.011111111111111112,
      "assigned_correctness_bounds": [
        0.011111111111111112,
        0.011111111111111112
      ],
      "wrong_solved_claims": 264,
      "eligible_solved_claims": 266,
      "wrong_claim_fraction": 0.9924812030075187,
      "wrong_controller_acceptance": 264,
      "actual_final_checks": 21,
      "mean_calls": 1.3222222222222222,
      "mean_tokens": 718.3416666666667,
      "mean_wall_seconds": 0.5107555555541896,
      "mean_api_seconds": 0.5017805555547562
    },
    {
      "model": "qwen2.5:0.5b",
      "condition": "single-contract",
      "assigned": 360,
      "observed": 359,
      "correct": 1,
      "success_fraction_observed": 0.002785515320334262,
      "assigned_correctness_bounds": [
        0.002777777777777778,
        0.005555555555555556
      ],
      "wrong_solved_claims": 867,
      "eligible_solved_claims": 867,
      "wrong_claim_fraction": 1.0,
      "wrong_controller_acceptance": 12,
      "actual_final_checks": 13,
      "mean_calls": 3.3091922005571033,
      "mean_tokens": 2865.1699164345405,
      "mean_wall_seconds": 0.8483593314754762,
      "mean_api_seconds": 0.8190668523691018
    },
    {
      "model": "qwen2.5:0.5b",
      "condition": "staged-open",
      "assigned": 360,
      "observed": 359,
      "correct": 0,
      "success_fraction_observed": 0.0,
      "assigned_correctness_bounds": [
        0.0,
        0.002777777777777778
      ],
      "wrong_solved_claims": 42,
      "eligible_solved_claims": 42,
      "wrong_claim_fraction": 1.0,
      "wrong_controller_acceptance": 42,
      "actual_final_checks": 19,
      "mean_calls": 4,
      "mean_tokens": 2585.114206128134,
      "mean_wall_seconds": 1.2370389972161306,
      "mean_api_seconds": 1.208701949864613
    },
    {
      "model": "qwen2.5:0.5b",
      "condition": "staged-contract",
      "assigned": 360,
      "observed": 359,
      "correct": 0,
      "success_fraction_observed": 0.0,
      "assigned_correctness_bounds": [
        0.0,
        0.002777777777777778
      ],
      "wrong_solved_claims": 70,
      "eligible_solved_claims": 70,
      "wrong_claim_fraction": 1.0,
      "wrong_controller_acceptance": 0,
      "actual_final_checks": 8,
      "mean_calls": 4,
      "mean_tokens": 2669.025069637883,
      "mean_wall_seconds": 1.2556239554316744,
      "mean_api_seconds": 1.2274651810605324
    }
  ],
  "models": {
    "qwen2.5-coder:0.5b": {
      "primary_contract_contrast": {
        "estimate": null,
        "identification_bounds": [
          -0.002777777777777777,
          -0.0013888888888888883
        ],
        "percentile_95_envelope": [
          -0.008333333333333333,
          0.005555555555555556
        ],
        "interpretation": "Paired generator-conditional percentile endpoints across extreme missing-outcome bounds; NOT full model/template-population inference",
        "n_instances": 120,
        "missingness_amendment": true
      },
      "complete_case_contract_descriptive": -0.0028011204481792713,
      "complete_case_instances": 119,
      "orchestration_complete_case_descriptive": 0.005602240896358544,
      "interaction_complete_case_descriptive": 0.011204481792717089,
      "deduplicated_contract_descriptive": -0.0024305555555555565,
      "deduplication_scope": "Post hoc, equal-weight unique payloads within cells; cells equal-weight; incomplete groups omitted. Not primary inference.",
      "no_preflight_overlap_contract_descriptive": -0.00046296296296296303,
      "no_preflight_overlap_unique_complete_payloads": 109,
      "no_preflight_overlap_scope": "Post hoc, exact preflight overlaps excluded; unique payloads equal-weight within cells, cells equal-weight; incomplete groups omitted. Not primary inference or strict unseen-template holdout."
    },
    "qwen2.5:0.5b": {
      "primary_contract_contrast": {
        "estimate": null,
        "identification_bounds": [
          -0.005555555555555555,
          -0.0013888888888888887
        ],
        "percentile_95_envelope": [
          -0.00972222222222222,
          0.0027777777777777775
        ],
        "interpretation": "Paired generator-conditional percentile endpoints across extreme missing-outcome bounds; NOT full model/template-population inference",
        "n_instances": 120,
        "missingness_amendment": true
      },
      "complete_case_contract_descriptive": -0.0028490028490028487,
      "complete_case_instances": 117,
      "orchestration_complete_case_descriptive": -0.005698005698005697,
      "interaction_complete_case_descriptive": 0.005698005698005697,
      "deduplicated_contract_descriptive": -0.0023148148148148147,
      "deduplication_scope": "Post hoc, equal-weight unique payloads within cells; cells equal-weight; incomplete groups omitted. Not primary inference.",
      "no_preflight_overlap_contract_descriptive": -0.001736111111111111,
      "no_preflight_overlap_unique_complete_payloads": 107,
      "no_preflight_overlap_scope": "Post hoc, exact preflight overlaps excluded; unique payloads equal-weight within cells, cells equal-weight; incomplete groups omitted. Not primary inference or strict unseen-template holdout."
    }
  },
  "families": [
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "capacity",
      "condition": "single-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.15
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "capacity",
      "condition": "single-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.15
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "capacity",
      "condition": "staged-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.15
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "capacity",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.15
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "dependency",
      "condition": "single-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.39999999999999997
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "dependency",
      "condition": "single-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.38888888888888884
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "dependency",
      "condition": "staged-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.39999999999999997
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "dependency",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.39444444444444443
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "freshness",
      "condition": "single-open",
      "observed": 60,
      "correct": 3,
      "mean_partial": 0.2222222222222222
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "freshness",
      "condition": "single-contract",
      "observed": 60,
      "correct": 2,
      "mean_partial": 0.17777777777777776
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "freshness",
      "condition": "staged-open",
      "observed": 60,
      "correct": 4,
      "mean_partial": 0.25
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "freshness",
      "condition": "staged-contract",
      "observed": 59,
      "correct": 4,
      "mean_partial": 0.2768361581920904
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "ranking",
      "condition": "single-open",
      "observed": 60,
      "correct": 1,
      "mean_partial": 0.12777777777777777
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "ranking",
      "condition": "single-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.17222222222222222
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "ranking",
      "condition": "staged-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.14444444444444443
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "ranking",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.1611111111111111
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "recovery",
      "condition": "single-open",
      "observed": 60,
      "correct": 58,
      "mean_partial": 0.9916666666666667
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "recovery",
      "condition": "single-contract",
      "observed": 60,
      "correct": 57,
      "mean_partial": 0.9875
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "recovery",
      "condition": "staged-open",
      "observed": 60,
      "correct": 58,
      "mean_partial": 0.9875
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "recovery",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 59,
      "mean_partial": 0.9958333333333333
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "validation",
      "condition": "single-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "validation",
      "condition": "single-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "validation",
      "condition": "staged-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "family": "validation",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "capacity",
      "condition": "single-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.5722222222222222
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "capacity",
      "condition": "single-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.6
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "capacity",
      "condition": "staged-open",
      "observed": 59,
      "correct": 0,
      "mean_partial": 0.5988700564971751
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "capacity",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.6166666666666666
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "dependency",
      "condition": "single-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.611111111111111
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "dependency",
      "condition": "single-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.6666666666666666
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "dependency",
      "condition": "staged-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.6277777777777778
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "dependency",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.6499999999999999
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "freshness",
      "condition": "single-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "freshness",
      "condition": "single-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "freshness",
      "condition": "staged-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "freshness",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "ranking",
      "condition": "single-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.005555555555555555
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "ranking",
      "condition": "single-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.0
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "ranking",
      "condition": "staged-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.0
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "ranking",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.0
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "recovery",
      "condition": "single-open",
      "observed": 60,
      "correct": 4,
      "mean_partial": 0.3625
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "recovery",
      "condition": "single-contract",
      "observed": 59,
      "correct": 1,
      "mean_partial": 0.326271186440678
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "recovery",
      "condition": "staged-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3125
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "recovery",
      "condition": "staged-contract",
      "observed": 59,
      "correct": 0,
      "mean_partial": 0.3135593220338983
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "validation",
      "condition": "single-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "validation",
      "condition": "single-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "validation",
      "condition": "staged-open",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    },
    {
      "model": "qwen2.5:0.5b",
      "family": "validation",
      "condition": "staged-contract",
      "observed": 60,
      "correct": 0,
      "mean_partial": 0.3333333333333333
    }
  ],
  "levels": [
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L1",
      "condition": "single-open",
      "observed": 90,
      "correct": 19,
      "mean_partial": 0.45555555555555555
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L1",
      "condition": "single-contract",
      "observed": 90,
      "correct": 16,
      "mean_partial": 0.46388888888888885
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L1",
      "condition": "staged-open",
      "observed": 90,
      "correct": 18,
      "mean_partial": 0.47129629629629627
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L1",
      "condition": "staged-contract",
      "observed": 90,
      "correct": 18,
      "mean_partial": 0.47129629629629627
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L2",
      "condition": "single-open",
      "observed": 90,
      "correct": 14,
      "mean_partial": 0.3675925925925926
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L2",
      "condition": "single-contract",
      "observed": 90,
      "correct": 13,
      "mean_partial": 0.3648148148148148
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L2",
      "condition": "staged-open",
      "observed": 90,
      "correct": 15,
      "mean_partial": 0.36666666666666664
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L2",
      "condition": "staged-contract",
      "observed": 90,
      "correct": 15,
      "mean_partial": 0.3925925925925926
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L3",
      "condition": "single-open",
      "observed": 90,
      "correct": 15,
      "mean_partial": 0.3222222222222222
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L3",
      "condition": "single-contract",
      "observed": 90,
      "correct": 15,
      "mean_partial": 0.3148148148148148
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L3",
      "condition": "staged-open",
      "observed": 90,
      "correct": 14,
      "mean_partial": 0.32407407407407407
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L3",
      "condition": "staged-contract",
      "observed": 89,
      "correct": 15,
      "mean_partial": 0.33707865168539325
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L4",
      "condition": "single-open",
      "observed": 90,
      "correct": 14,
      "mean_partial": 0.33796296296296297
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L4",
      "condition": "single-contract",
      "observed": 90,
      "correct": 15,
      "mean_partial": 0.3296296296296296
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L4",
      "condition": "staged-open",
      "observed": 90,
      "correct": 15,
      "mean_partial": 0.34814814814814815
    },
    {
      "model": "qwen2.5-coder:0.5b",
      "provisional_level": "L4",
      "condition": "staged-contract",
      "observed": 90,
      "correct": 15,
      "mean_partial": 0.34074074074074073
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L1",
      "condition": "single-open",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.39814814814814814
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L1",
      "condition": "single-contract",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.40555555555555556
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L1",
      "condition": "staged-open",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.40185185185185185
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L1",
      "condition": "staged-contract",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.4129629629629629
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L2",
      "condition": "single-open",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.35648148148148145
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L2",
      "condition": "single-contract",
      "observed": 89,
      "correct": 0,
      "mean_partial": 0.35767790262172283
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L2",
      "condition": "staged-open",
      "observed": 89,
      "correct": 0,
      "mean_partial": 0.37172284644194753
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L2",
      "condition": "staged-contract",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.3675925925925926
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L3",
      "condition": "single-open",
      "observed": 90,
      "correct": 1,
      "mean_partial": 0.33518518518518514
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L3",
      "condition": "single-contract",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.3675925925925926
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L3",
      "condition": "staged-open",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.3453703703703703
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L3",
      "condition": "staged-contract",
      "observed": 89,
      "correct": 0,
      "mean_partial": 0.36142322097378277
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L4",
      "condition": "single-open",
      "observed": 90,
      "correct": 3,
      "mean_partial": 0.3888888888888889
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L4",
      "condition": "single-contract",
      "observed": 90,
      "correct": 1,
      "mean_partial": 0.3759259259259259
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L4",
      "condition": "staged-open",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.34907407407407404
    },
    {
      "model": "qwen2.5:0.5b",
      "provisional_level": "L4",
      "condition": "staged-contract",
      "observed": 90,
      "correct": 0,
      "mean_partial": 0.35648148148148145
    }
  ],
  "stop_reasons": {
    "solved_finish": 1141,
    "model_call_cap": 1598,
    "unsolved_finish": 137
  },
  "infrastructure_interrupted": 4,
  "all_assignments_attempted": true,
  "semantic_payloads": {
    "draws": 120,
    "distinct": 115,
    "exact_repeat_draws": 5,
    "post_freeze_audit": true
  },
  "exact_preflight_overlap": {
    "evaluation_draws": 6,
    "evaluation_ids": [
      "recovery-L1-evaluation-3",
      "recovery-L1-evaluation-4",
      "recovery-L3-evaluation-1",
      "recovery-L3-evaluation-3",
      "recovery-L4-evaluation-0",
      "recovery-L4-evaluation-2"
    ],
    "distinct_semantic_payloads": 5,
    "post_freeze": true,
    "scope": "Exact cross-split payload overlap, not near-duplicate or unseen-template analysis"
  },
  "actual_tokens_observed_complete_trials": 6155745,
  "actual_calls_observed_complete_trials": 8549,
  "tokens_interrupted_request": "Unknown; completed partial turns retained, full interrupted request not reported as measured"
}