{
  "corpus_fingerprint": "278-8bc1c4f8addd",
  "revision": "2026-10-17",
  "published_at": "2026-10-17T01:12:58Z",
  "grader_version": "2026-10-10",
  "rows": [
    {
      "track": "closed",
      "model": "claude-opus-5.5",
      "provider": "claude-code",
      "total": 276,
      "correct": 132,
      "incorrect": 25,
      "abstained": 119,
      "no_answer": 0,
      "errors": 0,
      "truncated": 0,
      "score": 82,
      "score_per_100": 29.71014492753623,
      "runs": 1,
      "reasoning_correct": 42,
      "reasoning_total": 43,
      "grounded_correct": 90,
      "grounded_total": 233,
      "grounded_wrong": 24,
      "reasoning_accuracy": 97.67441860465117,
      "grounded_accuracy": 38.62660944206009,
      "grounded_hallucination_pct": 21.05263157894737,
      "cost_usd": 4.491024000000002
    },
    {
      "track": "closed",
      "model": "claude-fable-5.1",
      "provider": "claude-code",
      "total": 276,
      "correct": 139,
      "incorrect": 30,
      "abstained": 107,
      "no_answer": 0,
      "errors": 0,
      "truncated": 0,
      "score": 79,
      "score_per_100": 28.6231884057971,
      "runs": 1,
      "reasoning_correct": 43,
      "reasoning_total": 43,
      "grounded_correct": 96,
      "grounded_total": 233,
      "grounded_wrong": 30,
      "reasoning_accuracy": 100,
      "grounded_accuracy": 41.201716738197426,
      "grounded_hallucination_pct": 23.80952380952381,
      "cost_usd": 13.56906249999999
    },
    {
      "track": "closed",
      "model": "grok-4.7",
      "provider": "xai",
      "total": 276,
      "correct": 85,
      "incorrect": 5,
      "abstained": 186,
      "no_answer": 0,
      "errors": 0,
      "truncated": 0,
      "score": 75,
      "score_per_100": 27.17391304347826,
      "runs": 1,
      "reasoning_correct": 39,
      "reasoning_total": 43,
      "grounded_correct": 46,
      "grounded_total": 233,
      "grounded_wrong": 2,
      "reasoning_accuracy": 90.69767441860465,
      "grounded_accuracy": 19.742489270386265,
      "grounded_hallucination_pct": 4.166666666666667,
      "cost_usd": 2.978921999999998
    },
    {
      "track": "closed",
      "model": "claude-sonnet-5.5",
      "provider": "claude-code",
      "total": 276,
      "correct": 128,
      "incorrect": 45,
      "abstained": 103,
      "no_answer": 0,
      "errors": 0,
      "truncated": 0,
      "score": 38,
      "score_per_100": 13.768115942028986,
      "runs": 1,
      "reasoning_correct": 43,
      "reasoning_total": 43,
      "grounded_correct": 85,
      "grounded_total": 233,
      "grounded_wrong": 45,
      "reasoning_accuracy": 100,
      "grounded_accuracy": 36.48068669527897,
      "grounded_hallucination_pct": 34.61538461538461,
      "cost_usd": 2.2146999999999992
    },
    {
      "track": "closed",
      "model": "claude-haiku-5.5",
      "provider": "claude-code",
      "total": 276,
      "correct": 77,
      "incorrect": 32,
      "abstained": 167,
      "no_answer": 0,
      "errors": 0,
      "truncated": 0,
      "score": 13,
      "score_per_100": 4.7101449275362315,
      "runs": 1,
      "reasoning_correct": 41,
      "reasoning_total": 43,
      "grounded_correct": 36,
      "grounded_total": 233,
      "grounded_wrong": 30,
      "reasoning_accuracy": 95.34883720930233,
      "grounded_accuracy": 15.450643776824034,
      "grounded_hallucination_pct": 45.45454545454545,
      "cost_usd": 0.1267466000000001
    },
    {
      "track": "closed",
      "model": "gemma4-12b",
      "provider": "ollama",
      "total": 276,
      "correct": 30,
      "incorrect": 102,
      "abstained": 105,
      "no_answer": 0,
      "errors": 0,
      "truncated": 39,
      "score": -174,
      "score_per_100": -73.41772151898734,
      "runs": 1,
      "reasoning_correct": 22,
      "reasoning_total": 30,
      "grounded_correct": 8,
      "grounded_total": 207,
      "grounded_wrong": 94,
      "reasoning_accuracy": 73.33333333333333,
      "grounded_accuracy": 3.864734299516908,
      "grounded_hallucination_pct": 92.15686274509804,
      "cost_usd": 0
    },
    {
      "track": "closed",
      "model": "gpt-oss-20b",
      "provider": "ollama",
      "total": 276,
      "correct": 40,
      "incorrect": 145,
      "abstained": 88,
      "no_answer": 2,
      "errors": 0,
      "truncated": 1,
      "score": -250,
      "score_per_100": -90.9090909090909,
      "runs": 1,
      "reasoning_correct": 27,
      "reasoning_total": 40,
      "grounded_correct": 13,
      "grounded_total": 233,
      "grounded_wrong": 132,
      "reasoning_accuracy": 67.5,
      "grounded_accuracy": 5.579399141630901,
      "grounded_hallucination_pct": 91.03448275862068,
      "cost_usd": 0
    },
    {
      "track": "closed",
      "model": "deepseek-r1-14b",
      "provider": "ollama",
      "total": 276,
      "correct": 34,
      "incorrect": 145,
      "abstained": 96,
      "no_answer": 0,
      "errors": 0,
      "truncated": 1,
      "score": -256,
      "score_per_100": -93.0909090909091,
      "runs": 1,
      "reasoning_correct": 22,
      "reasoning_total": 42,
      "grounded_correct": 12,
      "grounded_total": 233,
      "grounded_wrong": 125,
      "reasoning_accuracy": 52.38095238095238,
      "grounded_accuracy": 5.150214592274678,
      "grounded_hallucination_pct": 91.24087591240875,
      "cost_usd": 0
    },
    {
      "track": "mcp",
      "model": "claude-opus-5.5",
      "provider": "claude-code",
      "total": 828,
      "correct": 717,
      "incorrect": 13,
      "abstained": 37,
      "no_answer": 61,
      "errors": 0,
      "truncated": 0,
      "score": 691,
      "score_per_100": 83.45410628019323,
      "runs": 3,
      "score_per_100_min": 82.2463768115942,
      "score_per_100_max": 84.78260869565217,
      "reasoning_correct": 126,
      "reasoning_total": 129,
      "grounded_correct": 591,
      "grounded_total": 638,
      "grounded_wrong": 10,
      "reasoning_accuracy": 97.67441860465117,
      "grounded_accuracy": 92.6332288401254,
      "grounded_hallucination_pct": 1.663893510815308,
      "cost_usd": 109.30653860000002
    },
    {
      "track": "mcp",
      "model": "claude-fable-5.1",
      "provider": "claude-code",
      "total": 552,
      "correct": 489,
      "incorrect": 18,
      "abstained": 31,
      "no_answer": 14,
      "errors": 0,
      "truncated": 0,
      "score": 453,
      "score_per_100": 82.06521739130434,
      "runs": 2,
      "score_per_100_min": 81.8840579710145,
      "score_per_100_max": 82.2463768115942,
      "reasoning_correct": 84,
      "reasoning_total": 86,
      "grounded_correct": 405,
      "grounded_total": 452,
      "grounded_wrong": 16,
      "reasoning_accuracy": 97.67441860465117,
      "grounded_accuracy": 89.60176991150442,
      "grounded_hallucination_pct": 3.8004750593824226,
      "cost_usd": 305.61237250000005
    },
    {
      "track": "mcp",
      "model": "claude-sonnet-5.5",
      "provider": "claude-code",
      "total": 828,
      "correct": 699,
      "incorrect": 48,
      "abstained": 47,
      "no_answer": 34,
      "errors": 0,
      "truncated": 0,
      "score": 603,
      "score_per_100": 72.82608695652173,
      "runs": 3,
      "score_per_100_min": 70.65217391304348,
      "score_per_100_max": 74.27536231884058,
      "reasoning_correct": 121,
      "reasoning_total": 129,
      "grounded_correct": 578,
      "grounded_total": 665,
      "grounded_wrong": 40,
      "reasoning_accuracy": 93.7984496124031,
      "grounded_accuracy": 86.9172932330827,
      "grounded_hallucination_pct": 6.472491909385114,
      "cost_usd": 43.64014809999996
    },
    {
      "track": "mcp",
      "model": "grok-4.7",
      "provider": "xai",
      "total": 552,
      "correct": 436,
      "incorrect": 21,
      "abstained": 95,
      "no_answer": 0,
      "errors": 0,
      "truncated": 0,
      "score": 394,
      "score_per_100": 71.3768115942029,
      "runs": 2,
      "score_per_100_min": 70.65217391304348,
      "score_per_100_max": 72.10144927536231,
      "reasoning_correct": 80,
      "reasoning_total": 86,
      "grounded_correct": 356,
      "grounded_total": 466,
      "grounded_wrong": 17,
      "reasoning_accuracy": 93.02325581395348,
      "grounded_accuracy": 76.39484978540773,
      "grounded_hallucination_pct": 4.557640750670242,
      "cost_usd": 54.312758000000045
    },
    {
      "track": "mcp",
      "model": "claude-haiku-5.5",
      "provider": "claude-code",
      "total": 828,
      "correct": 629,
      "incorrect": 55,
      "abstained": 99,
      "no_answer": 45,
      "errors": 0,
      "truncated": 0,
      "score": 519,
      "score_per_100": 62.68115942028985,
      "runs": 3,
      "score_per_100_min": 60.507246376811594,
      "score_per_100_max": 65.21739130434783,
      "reasoning_correct": 119,
      "reasoning_total": 129,
      "grounded_correct": 510,
      "grounded_total": 654,
      "grounded_wrong": 48,
      "reasoning_accuracy": 92.24806201550388,
      "grounded_accuracy": 77.98165137614679,
      "grounded_hallucination_pct": 8.602150537634408,
      "cost_usd": 4.191941170000002
    },
    {
      "track": "mcp",
      "model": "gpt-oss-20b",
      "provider": "ollama",
      "total": 828,
      "correct": 383,
      "incorrect": 173,
      "abstained": 44,
      "no_answer": 228,
      "errors": 0,
      "truncated": 0,
      "score": 37,
      "score_per_100": 4.468599033816425,
      "runs": 3,
      "score_per_100_min": 2.536231884057971,
      "score_per_100_max": 5.797101449275362,
      "reasoning_correct": 79,
      "reasoning_total": 107,
      "grounded_correct": 304,
      "grounded_total": 493,
      "grounded_wrong": 146,
      "reasoning_accuracy": 73.83177570093459,
      "grounded_accuracy": 61.66328600405679,
      "grounded_hallucination_pct": 32.44444444444444,
      "cost_usd": 0
    },
    {
      "track": "mcp",
      "model": "gemma4-12b",
      "provider": "ollama",
      "total": 828,
      "correct": 400,
      "incorrect": 214,
      "abstained": 186,
      "no_answer": 2,
      "errors": 0,
      "truncated": 26,
      "score": -28,
      "score_per_100": -3.491271820448878,
      "runs": 3,
      "score_per_100_min": -6.7164179104477615,
      "score_per_100_max": 1.8656716417910448,
      "reasoning_correct": 83,
      "reasoning_total": 118,
      "grounded_correct": 317,
      "grounded_total": 682,
      "grounded_wrong": 192,
      "reasoning_accuracy": 70.33898305084746,
      "grounded_accuracy": 46.48093841642229,
      "grounded_hallucination_pct": 37.721021611001966,
      "cost_usd": 0
    }
  ],
  "gaps": [
    {
      "model": "claude-fable-5.1",
      "score_gap_per_100": 53.44202898550724,
      "grounded_acc_closed": 41.201716738197426,
      "grounded_acc_mcp": 89.60176991150442,
      "halluc_closed": 23.80952380952381,
      "halluc_mcp": 3.8004750593824226
    },
    {
      "model": "claude-haiku-5.5",
      "score_gap_per_100": 57.97101449275362,
      "grounded_acc_closed": 15.450643776824034,
      "grounded_acc_mcp": 77.98165137614679,
      "halluc_closed": 45.45454545454545,
      "halluc_mcp": 8.602150537634408
    },
    {
      "model": "claude-opus-5.5",
      "score_gap_per_100": 53.743961352656996,
      "grounded_acc_closed": 38.62660944206009,
      "grounded_acc_mcp": 92.6332288401254,
      "halluc_closed": 21.05263157894737,
      "halluc_mcp": 1.663893510815308
    },
    {
      "model": "claude-sonnet-5.5",
      "score_gap_per_100": 59.05797101449275,
      "grounded_acc_closed": 36.48068669527897,
      "grounded_acc_mcp": 86.9172932330827,
      "halluc_closed": 34.61538461538461,
      "halluc_mcp": 6.472491909385114
    },
    {
      "model": "gemma4-12b",
      "score_gap_per_100": 69.92644969853846,
      "grounded_acc_closed": 3.864734299516908,
      "grounded_acc_mcp": 46.48093841642229,
      "halluc_closed": 92.15686274509804,
      "halluc_mcp": 37.721021611001966
    },
    {
      "model": "gpt-oss-20b",
      "score_gap_per_100": 95.37768994290734,
      "grounded_acc_closed": 5.579399141630901,
      "grounded_acc_mcp": 61.66328600405679,
      "halluc_closed": 91.03448275862068,
      "halluc_mcp": 32.44444444444444
    },
    {
      "model": "grok-4.7",
      "score_gap_per_100": 44.20289855072464,
      "grounded_acc_closed": 19.742489270386265,
      "grounded_acc_mcp": 76.39484978540773,
      "halluc_closed": 4.166666666666667,
      "halluc_mcp": 4.557640750670242
    }
  ],
  "jobs": [
    "baseline",
    "catalog",
    "constellation",
    "debris",
    "ground",
    "market",
    "regulatory"
  ],
  "errata": [
    {
      "id": "con-galileo-launched",
      "action": "rekey",
      "value": "34",
      "reason": "The question counts IOV and FOC satellites; the ground-truth SQL matched only names starting \"GALILEO \" and missed the 4 IOV satellites, giving 30. 34 Galileo satellites had been launched (latest 2025-12-17).",
      "date": "2026-10-11"
    },
    {
      "id": "deb-atlas5-centaur-afspc11-debris",
      "action": "drop",
      "reason": "The question names the wrong launch. The Centaur that fragmented in September 2024 is 2018-079B from the AEHF-4 launch (2018-10-17, 214 debris cataloged). AFSPC-11 (2018-036) has 1 debris object, which is what models answered.",
      "date": "2026-10-11"
    },
    {
      "id": "reg-ngso-market-access-count",
      "action": "rekey",
      "value": "9",
      "reason": "The question asks for foreign-licensed NGSO systems with US market access; the ground-truth SQL counted every granted NGSO system, U.S.-licensed ones included, giving 52. Excluding a U.S. licensing administration leaves 9 (Kepler, Telesat Ka-band, ExactView-1, ExactView-7, Kineis, Arctic Satellite Broadband Mission, two OneWeb systems, O3B).",
      "date": "2026-10-11"
    },
    {
      "id": "reg-noaa-remote-sensing-licensees",
      "action": "drop",
      "reason": "The question asks how many companies currently hold a license, but the answer key comes from a NOAA CRSRA licensee snapshot last seen 2024-07-27; no current list was available to verify it. To be restated as an as-of question, or rekeyed from a current list, in the next corpus.",
      "date": "2026-10-11"
    }
  ],
  "verdicts": [
    {
      "model": "claude-fable-5.1",
      "track": "closed",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 25,
      "gradeable": 25,
      "correct": 25,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.8668035060468213,
      "correct_hi": 1,
      "wrong_lo": 0,
      "wrong_hi": 0.13319649395317873,
      "verdict": "trust"
    },
    {
      "model": "claude-fable-5.1",
      "track": "mcp",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 2,
      "questions": 25,
      "gradeable": 50,
      "correct": 50,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.8668035060468213,
      "correct_hi": 1,
      "wrong_lo": 0,
      "wrong_hi": 0.13319649395317873,
      "verdict": "trust"
    },
    {
      "model": "claude-fable-5.1",
      "track": "closed",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 45,
      "gradeable": 45,
      "correct": 29,
      "incorrect": 3,
      "abstained": 13,
      "no_answer": 0,
      "correct_pct": 0.6444444444444445,
      "wrong_pct": 0.06666666666666667,
      "correct_lo": 0.49835546011952203,
      "correct_hi": 0.7678110864309596,
      "wrong_lo": 0.0229316164621471,
      "wrong_hi": 0.17856874388640825,
      "verdict": "verify"
    },
    {
      "model": "claude-fable-5.1",
      "track": "mcp",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 2,
      "questions": 45,
      "gradeable": 90,
      "correct": 89,
      "incorrect": 1,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 0.9888888888888889,
      "wrong_pct": 0.011111111111111112,
      "correct_lo": 0.9020324502033684,
      "correct_hi": 0.9988389381213386,
      "wrong_lo": 0.0011610618786614471,
      "wrong_hi": 0.09796754979663178,
      "verdict": "trust"
    },
    {
      "model": "claude-fable-5.1",
      "track": "closed",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 43,
      "gradeable": 43,
      "correct": 23,
      "incorrect": 9,
      "abstained": 11,
      "no_answer": 0,
      "correct_pct": 0.5348837209302325,
      "wrong_pct": 0.20930232558139536,
      "correct_lo": 0.38915389728872773,
      "correct_hi": 0.6748917373608144,
      "wrong_lo": 0.11422705997811457,
      "wrong_hi": 0.35205931794236645,
      "verdict": "dont_use"
    },
    {
      "model": "claude-fable-5.1",
      "track": "mcp",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 2,
      "questions": 43,
      "gradeable": 86,
      "correct": 86,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.9179874299767729,
      "correct_hi": 1,
      "wrong_lo": 0,
      "wrong_hi": 0.08201257002322722,
      "verdict": "trust"
    },
    {
      "model": "claude-fable-5.1",
      "track": "closed",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 42,
      "gradeable": 42,
      "correct": 19,
      "incorrect": 4,
      "abstained": 19,
      "no_answer": 0,
      "correct_pct": 0.4523809523809524,
      "wrong_pct": 0.09523809523809523,
      "correct_lo": 0.3122316381863852,
      "correct_hi": 0.6005113725287732,
      "wrong_lo": 0.03766163882831211,
      "wrong_hi": 0.22065395225053333,
      "verdict": "verify"
    },
    {
      "model": "claude-fable-5.1",
      "track": "mcp",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 2,
      "questions": 42,
      "gradeable": 84,
      "correct": 79,
      "incorrect": 2,
      "abstained": 3,
      "no_answer": 0,
      "correct_pct": 0.9404761904761905,
      "wrong_pct": 0.023809523809523808,
      "correct_lo": 0.8257573017536823,
      "correct_hi": 0.9813698491311036,
      "wrong_lo": 0.004215295499100304,
      "wrong_hi": 0.12321481165248255,
      "verdict": "trust"
    },
    {
      "model": "claude-fable-5.1",
      "track": "closed",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 22,
      "gradeable": 22,
      "correct": 6,
      "incorrect": 5,
      "abstained": 11,
      "no_answer": 0,
      "correct_pct": 0.2727272727272727,
      "wrong_pct": 0.22727272727272727,
      "correct_lo": 0.13150582015814438,
      "correct_hi": 0.4815212369900198,
      "wrong_lo": 0.10122890794820358,
      "wrong_hi": 0.4344035606295935,
      "verdict": "dont_use"
    },
    {
      "model": "claude-fable-5.1",
      "track": "mcp",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 2,
      "questions": 22,
      "gradeable": 44,
      "correct": 32,
      "incorrect": 6,
      "abstained": 2,
      "no_answer": 4,
      "correct_pct": 0.7272727272727273,
      "wrong_pct": 0.13636363636363635,
      "correct_lo": 0.5184787630099802,
      "correct_hi": 0.8684941798418556,
      "wrong_lo": 0.04748917264236269,
      "wrong_hi": 0.3333541187947,
      "verdict": "verify"
    },
    {
      "model": "claude-fable-5.1",
      "track": "closed",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 41,
      "gradeable": 41,
      "correct": 18,
      "incorrect": 7,
      "abstained": 16,
      "no_answer": 0,
      "correct_pct": 0.43902439024390244,
      "wrong_pct": 0.17073170731707318,
      "correct_lo": 0.2988990477210983,
      "correct_hi": 0.5895973484824271,
      "wrong_lo": 0.08525143995273901,
      "wrong_hi": 0.3126290995462977,
      "verdict": "verify"
    },
    {
      "model": "claude-fable-5.1",
      "track": "mcp",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 2,
      "questions": 41,
      "gradeable": 82,
      "correct": 74,
      "incorrect": 2,
      "abstained": 1,
      "no_answer": 5,
      "correct_pct": 0.9024390243902439,
      "wrong_pct": 0.024390243902439025,
      "correct_lo": 0.7745202448096945,
      "correct_hi": 0.9614035402470386,
      "wrong_lo": 0.00431842027071648,
      "wrong_hi": 0.12595347011678087,
      "verdict": "trust"
    },
    {
      "model": "claude-fable-5.1",
      "track": "closed",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 40,
      "gradeable": 40,
      "correct": 14,
      "incorrect": 2,
      "abstained": 24,
      "no_answer": 0,
      "correct_pct": 0.35,
      "wrong_pct": 0.05,
      "correct_lo": 0.2213433317104915,
      "correct_hi": 0.5049440300646263,
      "wrong_lo": 0.013820377146101098,
      "wrong_hi": 0.165041708179252,
      "verdict": "verify"
    },
    {
      "model": "claude-fable-5.1",
      "track": "mcp",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 2,
      "questions": 40,
      "gradeable": 80,
      "correct": 47,
      "incorrect": 3,
      "abstained": 25,
      "no_answer": 5,
      "correct_pct": 0.5875,
      "wrong_pct": 0.0375,
      "correct_lo": 0.4339080619817175,
      "correct_hi": 0.7257576436494639,
      "wrong_lo": 0.008707641902894095,
      "wrong_hi": 0.1473450569037188,
      "verdict": "verify"
    },
    {
      "model": "claude-fable-5.1",
      "track": "closed",
      "kind": "os-derived",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 13,
      "gradeable": 13,
      "correct": 0,
      "incorrect": 0,
      "abstained": 13,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0,
      "correct_lo": 0,
      "correct_hi": 0.22810184305529166,
      "wrong_lo": 0,
      "wrong_hi": 0.22810184305529166,
      "verdict": "verify"
    },
    {
      "model": "claude-fable-5.1",
      "track": "mcp",
      "kind": "os-derived",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 2,
      "questions": 13,
      "gradeable": 26,
      "correct": 22,
      "incorrect": 4,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 0.8461538461538461,
      "wrong_pct": 0.15384615384615385,
      "correct_lo": 0.5776482949136899,
      "correct_hi": 0.9567427368172621,
      "wrong_lo": 0.043257263182737954,
      "wrong_hi": 0.42235170508631015,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "closed",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 25,
      "gradeable": 25,
      "correct": 25,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.8668035060468213,
      "correct_hi": 1,
      "wrong_lo": 0,
      "wrong_hi": 0.13319649395317873,
      "verdict": "trust"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "mcp",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 25,
      "gradeable": 75,
      "correct": 74,
      "incorrect": 1,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 0.9866666666666667,
      "wrong_pct": 0.013333333333333334,
      "correct_lo": 0.8446808865024286,
      "correct_hi": 0.9990078593831441,
      "wrong_lo": 0.0009921406168558883,
      "wrong_hi": 0.15531911349757144,
      "verdict": "trust"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "closed",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 45,
      "gradeable": 45,
      "correct": 13,
      "incorrect": 8,
      "abstained": 24,
      "no_answer": 0,
      "correct_pct": 0.28888888888888886,
      "wrong_pct": 0.17777777777777778,
      "correct_lo": 0.17729918031900166,
      "correct_hi": 0.4336881747226023,
      "wrong_lo": 0.0929427892585383,
      "wrong_hi": 0.3133010684365414,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "mcp",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 45,
      "gradeable": 135,
      "correct": 123,
      "incorrect": 2,
      "abstained": 1,
      "no_answer": 9,
      "correct_pct": 0.9111111111111111,
      "wrong_pct": 0.014814814814814815,
      "correct_lo": 0.792661488316343,
      "correct_hi": 0.9648894518657969,
      "wrong_lo": 0.0019442110036230614,
      "wrong_hi": 0.10400918391655428,
      "verdict": "trust"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "closed",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 43,
      "gradeable": 43,
      "correct": 10,
      "incorrect": 6,
      "abstained": 27,
      "no_answer": 0,
      "correct_pct": 0.23255813953488372,
      "wrong_pct": 0.13953488372093023,
      "correct_lo": 0.1315353846924483,
      "correct_hi": 0.37744808299439414,
      "wrong_lo": 0.06555634267713947,
      "wrong_hi": 0.27263876594425684,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "mcp",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 43,
      "gradeable": 129,
      "correct": 106,
      "incorrect": 16,
      "abstained": 5,
      "no_answer": 2,
      "correct_pct": 0.8217054263565892,
      "wrong_pct": 0.12403100775193798,
      "correct_lo": 0.6825766401181178,
      "correct_hi": 0.9080664349832167,
      "wrong_lo": 0.05556165364366676,
      "wrong_hi": 0.2541687284668219,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "closed",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 42,
      "gradeable": 42,
      "correct": 6,
      "incorrect": 6,
      "abstained": 30,
      "no_answer": 0,
      "correct_pct": 0.14285714285714285,
      "wrong_pct": 0.14285714285714285,
      "correct_lo": 0.06715875684189936,
      "correct_hi": 0.2784138235217878,
      "wrong_lo": 0.06715875684189936,
      "wrong_hi": 0.2784138235217878,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "mcp",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 42,
      "gradeable": 126,
      "correct": 110,
      "incorrect": 1,
      "abstained": 13,
      "no_answer": 2,
      "correct_pct": 0.873015873015873,
      "wrong_pct": 0.007936507936507936,
      "correct_lo": 0.740428746279078,
      "correct_hi": 0.9430843364521821,
      "wrong_lo": 0.0005903553473814785,
      "wrong_hi": 0.09775408870925414,
      "verdict": "trust"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "closed",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 22,
      "gradeable": 22,
      "correct": 1,
      "incorrect": 2,
      "abstained": 19,
      "no_answer": 0,
      "correct_pct": 0.045454545454545456,
      "wrong_pct": 0.09090909090909091,
      "correct_lo": 0.008069219301147859,
      "correct_hi": 0.21798489499518053,
      "wrong_lo": 0.025294896282837764,
      "wrong_hi": 0.27815380658385785,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "mcp",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 22,
      "gradeable": 66,
      "correct": 35,
      "incorrect": 9,
      "abstained": 15,
      "no_answer": 7,
      "correct_pct": 0.5303030303030303,
      "wrong_pct": 0.13636363636363635,
      "correct_lo": 0.3333178989198782,
      "correct_hi": 0.7182784934603665,
      "wrong_lo": 0.04748917264236269,
      "wrong_hi": 0.3333541187947,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "closed",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 41,
      "gradeable": 41,
      "correct": 9,
      "incorrect": 5,
      "abstained": 27,
      "no_answer": 0,
      "correct_pct": 0.21951219512195122,
      "wrong_pct": 0.12195121951219512,
      "correct_lo": 0.12003040155386346,
      "correct_hi": 0.36705302098235293,
      "wrong_lo": 0.053232579973408245,
      "wrong_hi": 0.25544507648844866,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "mcp",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 41,
      "gradeable": 123,
      "correct": 101,
      "incorrect": 2,
      "abstained": 6,
      "no_answer": 14,
      "correct_pct": 0.8211382113821138,
      "wrong_pct": 0.016260162601626018,
      "correct_lo": 0.6781300384243848,
      "correct_hi": 0.9091222749037154,
      "wrong_lo": 0.0021342021336350214,
      "wrong_hi": 0.11327054108099902,
      "verdict": "trust"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "closed",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 40,
      "gradeable": 40,
      "correct": 8,
      "incorrect": 5,
      "abstained": 27,
      "no_answer": 0,
      "correct_pct": 0.2,
      "wrong_pct": 0.125,
      "correct_lo": 0.10499864973166874,
      "correct_hi": 0.34757607381856676,
      "wrong_lo": 0.05459419773541284,
      "wrong_hi": 0.26112420670238146,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "mcp",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 40,
      "gradeable": 120,
      "correct": 43,
      "incorrect": 13,
      "abstained": 53,
      "no_answer": 11,
      "correct_pct": 0.35833333333333334,
      "wrong_pct": 0.10833333333333334,
      "correct_lo": 0.22826293910901654,
      "correct_hi": 0.5132306803452614,
      "wrong_lo": 0.04445863307797926,
      "wrong_hi": 0.24084725600149476,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "closed",
      "kind": "os-derived",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 13,
      "gradeable": 13,
      "correct": 0,
      "incorrect": 0,
      "abstained": 13,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0,
      "correct_lo": 0,
      "correct_hi": 0.22810184305529166,
      "wrong_lo": 0,
      "wrong_hi": 0.22810184305529166,
      "verdict": "verify"
    },
    {
      "model": "claude-haiku-5.5",
      "track": "mcp",
      "kind": "os-derived",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 13,
      "gradeable": 39,
      "correct": 26,
      "incorrect": 7,
      "abstained": 6,
      "no_answer": 0,
      "correct_pct": 0.6666666666666666,
      "wrong_pct": 0.1794871794871795,
      "correct_lo": 0.4003196931375318,
      "correct_hi": 0.8569796925107044,
      "wrong_lo": 0.05526981696044017,
      "wrong_hi": 0.44992367217756735,
      "verdict": "verify"
    },
    {
      "model": "claude-opus-5.5",
      "track": "closed",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 25,
      "gradeable": 25,
      "correct": 25,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.8668035060468213,
      "correct_hi": 1,
      "wrong_lo": 0,
      "wrong_hi": 0.13319649395317873,
      "verdict": "trust"
    },
    {
      "model": "claude-opus-5.5",
      "track": "mcp",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 25,
      "gradeable": 75,
      "correct": 75,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.8668035060468213,
      "correct_hi": 1,
      "wrong_lo": 0,
      "wrong_hi": 0.13319649395317873,
      "verdict": "trust"
    },
    {
      "model": "claude-opus-5.5",
      "track": "closed",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 45,
      "gradeable": 45,
      "correct": 30,
      "incorrect": 5,
      "abstained": 10,
      "no_answer": 0,
      "correct_pct": 0.6666666666666666,
      "wrong_pct": 0.1111111111111111,
      "correct_lo": 0.520702168092554,
      "correct_hi": 0.7864130779272324,
      "wrong_lo": 0.04840400437825558,
      "wrong_hi": 0.23499375490890945,
      "verdict": "verify"
    },
    {
      "model": "claude-opus-5.5",
      "track": "mcp",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 45,
      "gradeable": 135,
      "correct": 135,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.9213457380593594,
      "correct_hi": 1,
      "wrong_lo": 6.938893903907228e-18,
      "wrong_hi": 0.07865426194064076,
      "verdict": "trust"
    },
    {
      "model": "claude-opus-5.5",
      "track": "closed",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 43,
      "gradeable": 43,
      "correct": 22,
      "incorrect": 5,
      "abstained": 16,
      "no_answer": 0,
      "correct_pct": 0.5116279069767442,
      "wrong_pct": 0.11627906976744186,
      "correct_lo": 0.36752063766074083,
      "correct_hi": 0.6538279072224399,
      "wrong_lo": 0.050703529948620846,
      "wrong_hi": 0.24479448890641395,
      "verdict": "verify"
    },
    {
      "model": "claude-opus-5.5",
      "track": "mcp",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 43,
      "gradeable": 129,
      "correct": 129,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.9179874299767729,
      "correct_hi": 1,
      "wrong_lo": 0,
      "wrong_hi": 0.08201257002322722,
      "verdict": "trust"
    },
    {
      "model": "claude-opus-5.5",
      "track": "closed",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 42,
      "gradeable": 42,
      "correct": 17,
      "incorrect": 3,
      "abstained": 22,
      "no_answer": 0,
      "correct_pct": 0.40476190476190477,
      "wrong_pct": 0.07142857142857142,
      "correct_lo": 0.27042647671209086,
      "correct_hi": 0.5550595447182258,
      "wrong_lo": 0.024589963551172878,
      "wrong_hi": 0.19009713288525168,
      "verdict": "verify"
    },
    {
      "model": "claude-opus-5.5",
      "track": "mcp",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 42,
      "gradeable": 126,
      "correct": 117,
      "incorrect": 3,
      "abstained": 6,
      "no_answer": 0,
      "correct_pct": 0.9285714285714286,
      "wrong_pct": 0.023809523809523808,
      "correct_lo": 0.8099028671147483,
      "correct_hi": 0.9754100364488272,
      "wrong_lo": 0.004215295499100304,
      "wrong_hi": 0.12321481165248255,
      "verdict": "trust"
    },
    {
      "model": "claude-opus-5.5",
      "track": "closed",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 22,
      "gradeable": 22,
      "correct": 4,
      "incorrect": 5,
      "abstained": 13,
      "no_answer": 0,
      "correct_pct": 0.18181818181818182,
      "wrong_pct": 0.22727272727272727,
      "correct_lo": 0.07306766024105121,
      "correct_hi": 0.38517021976637866,
      "wrong_lo": 0.10122890794820358,
      "wrong_hi": 0.4344035606295935,
      "verdict": "dont_use"
    },
    {
      "model": "claude-opus-5.5",
      "track": "mcp",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 22,
      "gradeable": 66,
      "correct": 45,
      "incorrect": 3,
      "abstained": 2,
      "no_answer": 16,
      "correct_pct": 0.6818181818181818,
      "wrong_pct": 0.045454545454545456,
      "correct_lo": 0.47318222782839336,
      "correct_hi": 0.8363961264530752,
      "wrong_lo": 0.008069219301147859,
      "wrong_hi": 0.21798489499518053,
      "verdict": "verify"
    },
    {
      "model": "claude-opus-5.5",
      "track": "closed",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 41,
      "gradeable": 41,
      "correct": 15,
      "incorrect": 6,
      "abstained": 20,
      "no_answer": 0,
      "correct_pct": 0.36585365853658536,
      "wrong_pct": 0.14634146341463414,
      "correct_lo": 0.23589630416936816,
      "correct_hi": 0.5187957674783876,
      "wrong_lo": 0.06884154529455412,
      "wrong_hi": 0.28443755268589266,
      "verdict": "verify"
    },
    {
      "model": "claude-opus-5.5",
      "track": "mcp",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 41,
      "gradeable": 123,
      "correct": 103,
      "incorrect": 2,
      "abstained": 5,
      "no_answer": 13,
      "correct_pct": 0.8373983739837398,
      "wrong_pct": 0.016260162601626018,
      "correct_lo": 0.6966872591434594,
      "correct_hi": 0.9202993485303674,
      "wrong_lo": 0.0021342021336350214,
      "wrong_hi": 0.11327054108099902,
      "verdict": "trust"
    },
    {
      "model": "claude-opus-5.5",
      "track": "closed",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 40,
      "gradeable": 40,
      "correct": 14,
      "incorrect": 1,
      "abstained": 25,
      "no_answer": 0,
      "correct_pct": 0.35,
      "wrong_pct": 0.025,
      "correct_lo": 0.2213433317104915,
      "correct_hi": 0.5049440300646263,
      "wrong_lo": 0.0044267174077949195,
      "wrong_hi": 0.1288165948800778,
      "verdict": "verify"
    },
    {
      "model": "claude-opus-5.5",
      "track": "mcp",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 40,
      "gradeable": 120,
      "correct": 68,
      "incorrect": 2,
      "abstained": 22,
      "no_answer": 28,
      "correct_pct": 0.5666666666666667,
      "wrong_pct": 0.016666666666666666,
      "correct_lo": 0.4140230650400791,
      "correct_hi": 0.7076269963932018,
      "wrong_lo": 0.002187647166978106,
      "wrong_hi": 0.1158494074417345,
      "verdict": "verify"
    },
    {
      "model": "claude-opus-5.5",
      "track": "closed",
      "kind": "os-derived",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 13,
      "gradeable": 13,
      "correct": 0,
      "incorrect": 0,
      "abstained": 13,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0,
      "correct_lo": 0,
      "correct_hi": 0.22810184305529166,
      "wrong_lo": 0,
      "wrong_hi": 0.22810184305529166,
      "verdict": "verify"
    },
    {
      "model": "claude-opus-5.5",
      "track": "mcp",
      "kind": "os-derived",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 13,
      "gradeable": 39,
      "correct": 30,
      "incorrect": 3,
      "abstained": 2,
      "no_answer": 4,
      "correct_pct": 0.7692307692307693,
      "wrong_pct": 0.07692307692307693,
      "correct_lo": 0.49743126004873084,
      "correct_hi": 0.918206209075343,
      "wrong_lo": 0.013710066379696295,
      "wrong_hi": 0.3331453392824736,
      "verdict": "verify"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "closed",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 25,
      "gradeable": 25,
      "correct": 25,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.8668035060468213,
      "correct_hi": 1,
      "wrong_lo": 0,
      "wrong_hi": 0.13319649395317873,
      "verdict": "trust"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "mcp",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 25,
      "gradeable": 75,
      "correct": 75,
      "incorrect": 0,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 1,
      "wrong_pct": 0,
      "correct_lo": 0.8668035060468213,
      "correct_hi": 1,
      "wrong_lo": 0,
      "wrong_hi": 0.13319649395317873,
      "verdict": "trust"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "closed",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 45,
      "gradeable": 45,
      "correct": 26,
      "incorrect": 11,
      "abstained": 8,
      "no_answer": 0,
      "correct_pct": 0.5777777777777777,
      "wrong_pct": 0.24444444444444444,
      "correct_lo": 0.433005327959445,
      "correct_hi": 0.7103151201831219,
      "wrong_lo": 0.14235335025509283,
      "wrong_hi": 0.3867366058479013,
      "verdict": "dont_use"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "mcp",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 45,
      "gradeable": 135,
      "correct": 132,
      "incorrect": 3,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 0.9777777777777777,
      "wrong_pct": 0.022222222222222223,
      "correct_lo": 0.8843305361128119,
      "correct_hi": 0.9960665024772425,
      "wrong_lo": 0.003933497522757505,
      "wrong_hi": 0.11566946388718813,
      "verdict": "trust"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "closed",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 43,
      "gradeable": 43,
      "correct": 21,
      "incorrect": 5,
      "abstained": 17,
      "no_answer": 0,
      "correct_pct": 0.4883720930232558,
      "wrong_pct": 0.11627906976744186,
      "correct_lo": 0.34617209277756,
      "correct_hi": 0.6324793623392592,
      "wrong_lo": 0.050703529948620846,
      "wrong_hi": 0.24479448890641395,
      "verdict": "verify"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "mcp",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 43,
      "gradeable": 129,
      "correct": 126,
      "incorrect": 2,
      "abstained": 1,
      "no_answer": 0,
      "correct_pct": 0.9767441860465116,
      "wrong_pct": 0.015503875968992248,
      "correct_lo": 0.8794073213878312,
      "correct_hi": 0.9958830188225801,
      "wrong_lo": 0.002034781229843595,
      "wrong_hi": 0.1084425153042913,
      "verdict": "trust"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "closed",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 42,
      "gradeable": 42,
      "correct": 19,
      "incorrect": 7,
      "abstained": 16,
      "no_answer": 0,
      "correct_pct": 0.4523809523809524,
      "wrong_pct": 0.16666666666666666,
      "correct_lo": 0.3122316381863852,
      "correct_hi": 0.6005113725287732,
      "wrong_lo": 0.08315811295836502,
      "wrong_hi": 0.30604296204774295,
      "verdict": "verify"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "mcp",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 42,
      "gradeable": 126,
      "correct": 117,
      "incorrect": 5,
      "abstained": 3,
      "no_answer": 1,
      "correct_pct": 0.9285714285714286,
      "wrong_pct": 0.03968253968253968,
      "correct_lo": 0.8099028671147483,
      "correct_hi": 0.9754100364488272,
      "wrong_lo": 0.009835992727494747,
      "wrong_hi": 0.14667977751903535,
      "verdict": "trust"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "closed",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 22,
      "gradeable": 22,
      "correct": 4,
      "incorrect": 6,
      "abstained": 12,
      "no_answer": 0,
      "correct_pct": 0.18181818181818182,
      "wrong_pct": 0.2727272727272727,
      "correct_lo": 0.07306766024105121,
      "correct_hi": 0.38517021976637866,
      "wrong_lo": 0.13150582015814438,
      "wrong_hi": 0.4815212369900198,
      "verdict": "dont_use"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "mcp",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 22,
      "gradeable": 66,
      "correct": 49,
      "incorrect": 11,
      "abstained": 3,
      "no_answer": 3,
      "correct_pct": 0.7424242424242424,
      "wrong_pct": 0.16666666666666666,
      "correct_lo": 0.5339703167166809,
      "correct_hi": 0.8788008223252772,
      "wrong_lo": 0.06422420536802462,
      "wrong_hi": 0.36821547844928293,
      "verdict": "verify"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "closed",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 41,
      "gradeable": 41,
      "correct": 17,
      "incorrect": 11,
      "abstained": 13,
      "no_answer": 0,
      "correct_pct": 0.4146341463414634,
      "wrong_pct": 0.2682926829268293,
      "correct_lo": 0.2775633082533015,
      "correct_hi": 0.5663316464316339,
      "wrong_lo": 0.15694805281336685,
      "wrong_hi": 0.41933825276002934,
      "verdict": "dont_use"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "mcp",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 41,
      "gradeable": 123,
      "correct": 102,
      "incorrect": 3,
      "abstained": 5,
      "no_answer": 13,
      "correct_pct": 0.8292682926829268,
      "wrong_pct": 0.024390243902439025,
      "correct_lo": 0.6873709004537024,
      "correct_hi": 0.914748560047261,
      "wrong_lo": 0.00431842027071648,
      "wrong_hi": 0.12595347011678087,
      "verdict": "trust"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "closed",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 40,
      "gradeable": 40,
      "correct": 11,
      "incorrect": 4,
      "abstained": 25,
      "no_answer": 0,
      "correct_pct": 0.275,
      "wrong_pct": 0.1,
      "correct_lo": 0.1610785343314309,
      "correct_hi": 0.4283525083312457,
      "wrong_lo": 0.03957888759557489,
      "wrong_hi": 0.23052074380473903,
      "verdict": "verify"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "mcp",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 40,
      "gradeable": 120,
      "correct": 56,
      "incorrect": 17,
      "abstained": 34,
      "no_answer": 13,
      "correct_pct": 0.4666666666666667,
      "wrong_pct": 0.14166666666666666,
      "correct_lo": 0.32188066221605804,
      "correct_hi": 0.6172943070673015,
      "wrong_lo": 0.06517299126498999,
      "wrong_hi": 0.28095792853112456,
      "verdict": "verify"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "closed",
      "kind": "os-derived",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 1,
      "questions": 13,
      "gradeable": 13,
      "correct": 0,
      "incorrect": 1,
      "abstained": 12,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0.07692307692307693,
      "correct_lo": 0,
      "correct_hi": 0.22810184305529166,
      "wrong_lo": 0.013710066379696295,
      "wrong_hi": 0.3331453392824736,
      "verdict": "verify"
    },
    {
      "model": "claude-sonnet-5.5",
      "track": "mcp",
      "kind": "os-derived",
      "provider": "claude-code",
      "harness": "2.1.296 (Claude Code)",
      "runs": 3,
      "questions": 13,
      "gradeable": 39,
      "correct": 30,
      "incorrect": 4,
      "abstained": 1,
      "no_answer": 4,
      "correct_pct": 0.7692307692307693,
      "wrong_pct": 0.10256410256410256,
      "correct_lo": 0.49743126004873084,
      "correct_hi": 0.918206209075343,
      "wrong_lo": 0.022298775323178394,
      "wrong_hi": 0.3641411512079509,
      "verdict": "verify"
    },
    {
      "model": "deepseek-r1-14b",
      "track": "closed",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 24,
      "gradeable": 24,
      "correct": 14,
      "incorrect": 10,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 0.5833333333333334,
      "wrong_pct": 0.4166666666666667,
      "correct_lo": 0.3883433764938864,
      "correct_hi": 0.7553265275346177,
      "wrong_lo": 0.24467347246538232,
      "wrong_hi": 0.6116566235061137,
      "verdict": "dont_use"
    },
    {
      "model": "deepseek-r1-14b",
      "track": "closed",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 45,
      "gradeable": 45,
      "correct": 1,
      "incorrect": 34,
      "abstained": 10,
      "no_answer": 0,
      "correct_pct": 0.022222222222222223,
      "wrong_pct": 0.7555555555555555,
      "correct_lo": 0.003933497522757505,
      "correct_hi": 0.11566946388718813,
      "wrong_lo": 0.6132633941520986,
      "wrong_hi": 0.8576466497449071,
      "verdict": "dont_use"
    },
    {
      "model": "deepseek-r1-14b",
      "track": "closed",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 43,
      "gradeable": 43,
      "correct": 3,
      "incorrect": 29,
      "abstained": 11,
      "no_answer": 0,
      "correct_pct": 0.06976744186046512,
      "wrong_pct": 0.6744186046511628,
      "correct_lo": 0.024011153791962503,
      "correct_hi": 0.18609268553034927,
      "wrong_lo": 0.5251593250655096,
      "wrong_hi": 0.7950688481822019,
      "verdict": "dont_use"
    },
    {
      "model": "deepseek-r1-14b",
      "track": "closed",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 42,
      "gradeable": 42,
      "correct": 5,
      "incorrect": 21,
      "abstained": 16,
      "no_answer": 0,
      "correct_pct": 0.11904761904761904,
      "wrong_pct": 0.5,
      "correct_lo": 0.05193726765019084,
      "correct_hi": 0.2500068180710754,
      "wrong_lo": 0.3552574591652803,
      "wrong_hi": 0.6447425408347197,
      "verdict": "dont_use"
    },
    {
      "model": "deepseek-r1-14b",
      "track": "closed",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 22,
      "gradeable": 22,
      "correct": 3,
      "incorrect": 8,
      "abstained": 11,
      "no_answer": 0,
      "correct_pct": 0.13636363636363635,
      "wrong_pct": 0.36363636363636365,
      "correct_lo": 0.04748917264236269,
      "correct_hi": 0.3333541187947,
      "wrong_lo": 0.19732972772607899,
      "wrong_hi": 0.5704865065628195,
      "verdict": "dont_use"
    },
    {
      "model": "deepseek-r1-14b",
      "track": "closed",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 41,
      "gradeable": 41,
      "correct": 1,
      "incorrect": 18,
      "abstained": 22,
      "no_answer": 0,
      "correct_pct": 0.024390243902439025,
      "wrong_pct": 0.43902439024390244,
      "correct_lo": 0.00431842027071648,
      "correct_hi": 0.12595347011678087,
      "wrong_lo": 0.2988990477210983,
      "wrong_hi": 0.5895973484824271,
      "verdict": "dont_use"
    },
    {
      "model": "deepseek-r1-14b",
      "track": "closed",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 40,
      "gradeable": 40,
      "correct": 3,
      "incorrect": 20,
      "abstained": 17,
      "no_answer": 0,
      "correct_pct": 0.075,
      "wrong_pct": 0.5,
      "correct_lo": 0.025835556771858226,
      "correct_hi": 0.19864530159097524,
      "wrong_lo": 0.35199278797099753,
      "wrong_hi": 0.6480072120290025,
      "verdict": "dont_use"
    },
    {
      "model": "deepseek-r1-14b",
      "track": "closed",
      "kind": "os-derived",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 13,
      "gradeable": 13,
      "correct": 0,
      "incorrect": 4,
      "abstained": 9,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0.3076923076923077,
      "correct_lo": 0,
      "correct_hi": 0.22810184305529166,
      "wrong_lo": 0.12680497161618087,
      "wrong_hi": 0.5763111218666236,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "closed",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 13,
      "gradeable": 13,
      "correct": 12,
      "incorrect": 1,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 0.9230769230769231,
      "wrong_pct": 0.07692307692307693,
      "correct_lo": 0.6668546607175265,
      "correct_hi": 0.9862899336203037,
      "wrong_lo": 0.013710066379696295,
      "wrong_hi": 0.3331453392824736,
      "verdict": "verify"
    },
    {
      "model": "gemma4-12b",
      "track": "mcp",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 23,
      "gradeable": 64,
      "correct": 57,
      "incorrect": 2,
      "abstained": 5,
      "no_answer": 0,
      "correct_pct": 0.890625,
      "wrong_pct": 0.03125,
      "correct_lo": 0.7040764768585807,
      "correct_hi": 0.9653601439091829,
      "wrong_lo": 0.00435094393543034,
      "wrong_hi": 0.1923251111432535,
      "verdict": "trust"
    },
    {
      "model": "gemma4-12b",
      "track": "closed",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 38,
      "gradeable": 38,
      "correct": 0,
      "incorrect": 26,
      "abstained": 12,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0.6842105263157895,
      "correct_lo": 0,
      "correct_hi": 0.09181293258384,
      "wrong_lo": 0.5254394356005149,
      "wrong_hi": 0.8091557997633335,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "mcp",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 45,
      "gradeable": 130,
      "correct": 101,
      "incorrect": 19,
      "abstained": 10,
      "no_answer": 0,
      "correct_pct": 0.7769230769230769,
      "wrong_pct": 0.14615384615384616,
      "correct_lo": 0.6363720610905004,
      "correct_hi": 0.8739117322962218,
      "wrong_lo": 0.07107720955275645,
      "wrong_hi": 0.2768934988975432,
      "verdict": "verify"
    },
    {
      "model": "gemma4-12b",
      "track": "closed",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 33,
      "gradeable": 33,
      "correct": 2,
      "incorrect": 19,
      "abstained": 12,
      "no_answer": 0,
      "correct_pct": 0.06060606060606061,
      "wrong_pct": 0.5757575757575758,
      "correct_lo": 0.01678053042907096,
      "correct_hi": 0.19606581717798194,
      "wrong_lo": 0.4080697369146003,
      "wrong_hi": 0.7276464100497011,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "mcp",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 43,
      "gradeable": 128,
      "correct": 54,
      "incorrect": 52,
      "abstained": 21,
      "no_answer": 1,
      "correct_pct": 0.421875,
      "wrong_pct": 0.40625,
      "correct_lo": 0.28670670404341114,
      "correct_hi": 0.5698577600227182,
      "wrong_lo": 0.27307901537662815,
      "wrong_hi": 0.554798341502727,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "closed",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 41,
      "gradeable": 41,
      "correct": 4,
      "incorrect": 14,
      "abstained": 23,
      "no_answer": 0,
      "correct_pct": 0.0975609756097561,
      "wrong_pct": 0.34146341463414637,
      "correct_lo": 0.03859645975296143,
      "correct_hi": 0.22547975519030555,
      "wrong_lo": 0.2155861736664346,
      "wrong_hi": 0.4945044564627312,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "mcp",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 42,
      "gradeable": 125,
      "correct": 65,
      "incorrect": 22,
      "abstained": 38,
      "no_answer": 0,
      "correct_pct": 0.52,
      "wrong_pct": 0.176,
      "correct_lo": 0.3736875561295388,
      "correct_hi": 0.6629603793700949,
      "wrong_lo": 0.08961555603586711,
      "wrong_hi": 0.3166878888700699,
      "verdict": "verify"
    },
    {
      "model": "gemma4-12b",
      "track": "closed",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 22,
      "gradeable": 22,
      "correct": 0,
      "incorrect": 10,
      "abstained": 12,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0.45454545454545453,
      "correct_lo": 0,
      "correct_hi": 0.14865952572596125,
      "wrong_lo": 0.26920014432004846,
      "wrong_hi": 0.6534052671095845,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "mcp",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 22,
      "gradeable": 64,
      "correct": 12,
      "incorrect": 37,
      "abstained": 14,
      "no_answer": 1,
      "correct_pct": 0.1875,
      "wrong_pct": 0.578125,
      "correct_lo": 0.0764584223912903,
      "correct_hi": 0.39145378118743546,
      "wrong_lo": 0.3757429969450212,
      "wrong_hi": 0.7572789521602974,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "closed",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 36,
      "gradeable": 36,
      "correct": 3,
      "incorrect": 11,
      "abstained": 22,
      "no_answer": 0,
      "correct_pct": 0.08333333333333333,
      "wrong_pct": 0.3055555555555556,
      "correct_lo": 0.0287480892712918,
      "correct_hi": 0.21827010276918848,
      "wrong_lo": 0.18004270282805354,
      "wrong_hi": 0.46856578679083716,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "mcp",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 41,
      "gradeable": 120,
      "correct": 56,
      "incorrect": 43,
      "abstained": 21,
      "no_answer": 0,
      "correct_pct": 0.4666666666666667,
      "wrong_pct": 0.35833333333333334,
      "correct_lo": 0.3234724330366314,
      "correct_hi": 0.6155722635546291,
      "wrong_lo": 0.22959573810529016,
      "wrong_hi": 0.511344222407567,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "closed",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 36,
      "gradeable": 36,
      "correct": 4,
      "incorrect": 16,
      "abstained": 16,
      "no_answer": 0,
      "correct_pct": 0.1111111111111111,
      "wrong_pct": 0.4444444444444444,
      "correct_lo": 0.04406568908741833,
      "correct_hi": 0.25315129015036325,
      "wrong_lo": 0.2954102987654314,
      "wrong_hi": 0.6041921268399659,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "mcp",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 40,
      "gradeable": 120,
      "correct": 16,
      "incorrect": 31,
      "abstained": 73,
      "no_answer": 0,
      "correct_pct": 0.13333333333333333,
      "wrong_pct": 0.25833333333333336,
      "correct_lo": 0.05983196228036035,
      "correct_hi": 0.27109269983659406,
      "wrong_lo": 0.14821936174138453,
      "wrong_hi": 0.4107991655629719,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "closed",
      "kind": "os-derived",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 13,
      "gradeable": 13,
      "correct": 0,
      "incorrect": 5,
      "abstained": 8,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0.38461538461538464,
      "correct_lo": 0,
      "correct_hi": 0.22810184305529166,
      "wrong_lo": 0.17709446581236818,
      "wrong_hi": 0.6447751902773146,
      "verdict": "dont_use"
    },
    {
      "model": "gemma4-12b",
      "track": "mcp",
      "kind": "os-derived",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 13,
      "gradeable": 36,
      "correct": 27,
      "incorrect": 5,
      "abstained": 4,
      "no_answer": 0,
      "correct_pct": 0.75,
      "wrong_pct": 0.1388888888888889,
      "correct_lo": 0.4784494853422473,
      "correct_hi": 0.907499593130107,
      "wrong_lo": 0.03669049170196942,
      "wrong_hi": 0.40582750606018575,
      "verdict": "verify"
    },
    {
      "model": "gpt-oss-20b",
      "track": "closed",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 24,
      "gradeable": 24,
      "correct": 18,
      "incorrect": 4,
      "abstained": 0,
      "no_answer": 2,
      "correct_pct": 0.75,
      "wrong_pct": 0.16666666666666666,
      "correct_lo": 0.5510017468789357,
      "correct_hi": 0.8800079652065766,
      "wrong_lo": 0.06678568000728921,
      "wrong_hi": 0.35853470387869435,
      "verdict": "verify"
    },
    {
      "model": "gpt-oss-20b",
      "track": "mcp",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 25,
      "gradeable": 75,
      "correct": 64,
      "incorrect": 5,
      "abstained": 0,
      "no_answer": 6,
      "correct_pct": 0.8533333333333334,
      "wrong_pct": 0.06666666666666667,
      "correct_lo": 0.6688471946120905,
      "correct_hi": 0.9436939496609965,
      "wrong_lo": 0.016592731775468883,
      "wrong_hi": 0.23217756298395267,
      "verdict": "verify"
    },
    {
      "model": "gpt-oss-20b",
      "track": "closed",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 45,
      "gradeable": 45,
      "correct": 3,
      "incorrect": 28,
      "abstained": 14,
      "no_answer": 0,
      "correct_pct": 0.06666666666666667,
      "wrong_pct": 0.6222222222222222,
      "correct_lo": 0.0229316164621471,
      "correct_hi": 0.17856874388640825,
      "wrong_lo": 0.4762967000646754,
      "wrong_hi": 0.7489211470165014,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "mcp",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 45,
      "gradeable": 135,
      "correct": 102,
      "incorrect": 14,
      "abstained": 2,
      "no_answer": 17,
      "correct_pct": 0.7555555555555555,
      "wrong_pct": 0.1037037037037037,
      "correct_lo": 0.6132633941520986,
      "correct_hi": 0.8576466497449071,
      "wrong_lo": 0.04386614242327462,
      "wrong_hi": 0.22588205037412212,
      "verdict": "verify"
    },
    {
      "model": "gpt-oss-20b",
      "track": "closed",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 43,
      "gradeable": 43,
      "correct": 2,
      "incorrect": 22,
      "abstained": 19,
      "no_answer": 0,
      "correct_pct": 0.046511627906976744,
      "wrong_pct": 0.5116279069767442,
      "correct_lo": 0.012848990317319689,
      "correct_hi": 0.15455775923863058,
      "wrong_lo": 0.36752063766074083,
      "wrong_hi": 0.6538279072224399,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "mcp",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 43,
      "gradeable": 129,
      "correct": 38,
      "incorrect": 47,
      "abstained": 4,
      "no_answer": 40,
      "correct_pct": 0.29457364341085274,
      "wrong_pct": 0.3643410852713178,
      "correct_lo": 0.17979272202216878,
      "correct_hi": 0.44304965170830446,
      "wrong_lo": 0.2372003425387799,
      "wrong_hi": 0.5137333004907779,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "closed",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 42,
      "gradeable": 42,
      "correct": 5,
      "incorrect": 27,
      "abstained": 10,
      "no_answer": 0,
      "correct_pct": 0.11904761904761904,
      "wrong_pct": 0.6428571428571429,
      "correct_lo": 0.05193726765019084,
      "correct_hi": 0.2500068180710754,
      "wrong_lo": 0.4916608928707938,
      "wrong_hi": 0.7701100749837315,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "mcp",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 42,
      "gradeable": 126,
      "correct": 56,
      "incorrect": 24,
      "abstained": 9,
      "no_answer": 37,
      "correct_pct": 0.4444444444444444,
      "wrong_pct": 0.19047619047619047,
      "correct_lo": 0.3051784747834041,
      "correct_hi": 0.5930217043842806,
      "wrong_lo": 0.09981889609856438,
      "wrong_hi": 0.3330106735499645,
      "verdict": "verify"
    },
    {
      "model": "gpt-oss-20b",
      "track": "closed",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 22,
      "gradeable": 22,
      "correct": 2,
      "incorrect": 10,
      "abstained": 10,
      "no_answer": 0,
      "correct_pct": 0.09090909090909091,
      "wrong_pct": 0.45454545454545453,
      "correct_lo": 0.025294896282837764,
      "correct_hi": 0.27815380658385785,
      "wrong_lo": 0.26920014432004846,
      "wrong_hi": 0.6534052671095845,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "mcp",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 22,
      "gradeable": 66,
      "correct": 15,
      "incorrect": 12,
      "abstained": 11,
      "no_answer": 28,
      "correct_pct": 0.22727272727272727,
      "wrong_pct": 0.18181818181818182,
      "correct_lo": 0.10122890794820358,
      "correct_hi": 0.4344035606295935,
      "wrong_lo": 0.07306766024105121,
      "wrong_hi": 0.38517021976637866,
      "verdict": "verify"
    },
    {
      "model": "gpt-oss-20b",
      "track": "closed",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 41,
      "gradeable": 41,
      "correct": 3,
      "incorrect": 24,
      "abstained": 14,
      "no_answer": 0,
      "correct_pct": 0.07317073170731707,
      "wrong_pct": 0.5853658536585366,
      "correct_lo": 0.02519737325256413,
      "correct_hi": 0.19427740017211295,
      "wrong_lo": 0.433668353568366,
      "wrong_hi": 0.7224366917466984,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "mcp",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 41,
      "gradeable": 123,
      "correct": 57,
      "incorrect": 26,
      "abstained": 6,
      "no_answer": 34,
      "correct_pct": 0.4634146341463415,
      "wrong_pct": 0.21138211382113822,
      "correct_lo": 0.32055999118185885,
      "correct_hi": 0.6125378465402564,
      "wrong_lo": 0.11407283240921819,
      "wrong_hi": 0.3581434429541349,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "closed",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 40,
      "gradeable": 40,
      "correct": 3,
      "incorrect": 24,
      "abstained": 13,
      "no_answer": 0,
      "correct_pct": 0.075,
      "wrong_pct": 0.6,
      "correct_lo": 0.025835556771858226,
      "correct_hi": 0.19864530159097524,
      "wrong_lo": 0.445956198273935,
      "wrong_hi": 0.7365188938759863,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "mcp",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 40,
      "gradeable": 120,
      "correct": 16,
      "incorrect": 30,
      "abstained": 10,
      "no_answer": 64,
      "correct_pct": 0.13333333333333333,
      "wrong_pct": 0.25,
      "correct_lo": 0.05983196228036035,
      "correct_hi": 0.27109269983659406,
      "wrong_lo": 0.14186967895549518,
      "wrong_hi": 0.40194259066970106,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "closed",
      "kind": "os-derived",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 13,
      "gradeable": 13,
      "correct": 0,
      "incorrect": 5,
      "abstained": 8,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0.38461538461538464,
      "correct_lo": 0,
      "correct_hi": 0.22810184305529166,
      "wrong_lo": 0.17709446581236818,
      "wrong_hi": 0.6447751902773146,
      "verdict": "dont_use"
    },
    {
      "model": "gpt-oss-20b",
      "track": "mcp",
      "kind": "os-derived",
      "provider": "ollama",
      "harness": "spacebench-runner",
      "runs": 3,
      "questions": 13,
      "gradeable": 39,
      "correct": 26,
      "incorrect": 11,
      "abstained": 2,
      "no_answer": 0,
      "correct_pct": 0.6666666666666666,
      "wrong_pct": 0.28205128205128205,
      "correct_lo": 0.4003196931375318,
      "correct_hi": 0.8569796925107044,
      "wrong_lo": 0.1111716335484369,
      "wrong_hi": 0.5523599390654081,
      "verdict": "dont_use"
    },
    {
      "model": "grok-4.7",
      "track": "closed",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 25,
      "gradeable": 25,
      "correct": 24,
      "incorrect": 0,
      "abstained": 1,
      "no_answer": 0,
      "correct_pct": 0.96,
      "wrong_pct": 0,
      "correct_lo": 0.8045552759497747,
      "correct_hi": 0.9929039496133008,
      "wrong_lo": 0,
      "wrong_hi": 0.13319649395317873,
      "verdict": "trust"
    },
    {
      "model": "grok-4.7",
      "track": "mcp",
      "job": "baseline",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 2,
      "questions": 25,
      "gradeable": 50,
      "correct": 49,
      "incorrect": 0,
      "abstained": 1,
      "no_answer": 0,
      "correct_pct": 0.98,
      "wrong_pct": 0,
      "correct_lo": 0.8342228568014084,
      "correct_hi": 0.9979085090035399,
      "wrong_lo": 0,
      "wrong_hi": 0.13319649395317873,
      "verdict": "trust"
    },
    {
      "model": "grok-4.7",
      "track": "closed",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 45,
      "gradeable": 45,
      "correct": 21,
      "incorrect": 0,
      "abstained": 24,
      "no_answer": 0,
      "correct_pct": 0.4666666666666667,
      "wrong_pct": 0,
      "correct_lo": 0.3293490394201473,
      "correct_hi": 0.6092279113758954,
      "wrong_lo": 6.938893903907228e-18,
      "wrong_hi": 0.07865426194064076,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "mcp",
      "job": "catalog",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 2,
      "questions": 45,
      "gradeable": 90,
      "correct": 82,
      "incorrect": 1,
      "abstained": 7,
      "no_answer": 0,
      "correct_pct": 0.9111111111111111,
      "wrong_pct": 0.011111111111111112,
      "correct_lo": 0.792661488316343,
      "correct_hi": 0.9648894518657969,
      "wrong_lo": 0.0011610618786614471,
      "wrong_hi": 0.09796754979663178,
      "verdict": "trust"
    },
    {
      "model": "grok-4.7",
      "track": "closed",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 43,
      "gradeable": 43,
      "correct": 8,
      "incorrect": 0,
      "abstained": 35,
      "no_answer": 0,
      "correct_pct": 0.18604651162790697,
      "wrong_pct": 0,
      "correct_lo": 0.0974163558010598,
      "correct_hi": 0.3261729323530596,
      "wrong_lo": 0,
      "wrong_hi": 0.08201257002322722,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "mcp",
      "job": "constellation",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 2,
      "questions": 43,
      "gradeable": 86,
      "correct": 82,
      "incorrect": 4,
      "abstained": 0,
      "no_answer": 0,
      "correct_pct": 0.9534883720930233,
      "wrong_pct": 0.046511627906976744,
      "correct_lo": 0.8454422407613695,
      "correct_hi": 0.9871510096826803,
      "wrong_lo": 0.012848990317319689,
      "wrong_hi": 0.15455775923863058,
      "verdict": "trust"
    },
    {
      "model": "grok-4.7",
      "track": "closed",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 42,
      "gradeable": 42,
      "correct": 5,
      "incorrect": 2,
      "abstained": 35,
      "no_answer": 0,
      "correct_pct": 0.11904761904761904,
      "wrong_pct": 0.047619047619047616,
      "correct_lo": 0.05193726765019084,
      "correct_hi": 0.2500068180710754,
      "wrong_lo": 0.013157248639206379,
      "wrong_hi": 0.1579013531547973,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "mcp",
      "job": "debris",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 2,
      "questions": 42,
      "gradeable": 84,
      "correct": 71,
      "incorrect": 3,
      "abstained": 10,
      "no_answer": 0,
      "correct_pct": 0.8452380952380952,
      "wrong_pct": 0.03571428571428571,
      "correct_lo": 0.7076825728932761,
      "correct_hi": 0.9249305994218264,
      "wrong_lo": 0.008290820071836139,
      "wrong_hi": 0.14095353440095715,
      "verdict": "trust"
    },
    {
      "model": "grok-4.7",
      "track": "closed",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 22,
      "gradeable": 22,
      "correct": 3,
      "incorrect": 1,
      "abstained": 18,
      "no_answer": 0,
      "correct_pct": 0.13636363636363635,
      "wrong_pct": 0.045454545454545456,
      "correct_lo": 0.04748917264236269,
      "correct_hi": 0.3333541187947,
      "wrong_lo": 0.008069219301147859,
      "wrong_hi": 0.21798489499518053,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "mcp",
      "job": "ground",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 2,
      "questions": 22,
      "gradeable": 44,
      "correct": 25,
      "incorrect": 5,
      "abstained": 14,
      "no_answer": 0,
      "correct_pct": 0.5681818181818182,
      "wrong_pct": 0.11363636363636363,
      "correct_lo": 0.36679601352424396,
      "correct_hi": 0.7492958693313069,
      "wrong_lo": 0.03589747812594399,
      "wrong_hi": 0.3062485190259352,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "closed",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 41,
      "gradeable": 41,
      "correct": 8,
      "incorrect": 2,
      "abstained": 31,
      "no_answer": 0,
      "correct_pct": 0.1951219512195122,
      "wrong_pct": 0.04878048780487805,
      "correct_lo": 0.10234315749205929,
      "correct_hi": 0.34013882352556724,
      "wrong_lo": 0.013480662282573552,
      "wrong_hi": 0.16139266962351367,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "mcp",
      "job": "market",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 2,
      "questions": 41,
      "gradeable": 82,
      "correct": 64,
      "incorrect": 2,
      "abstained": 16,
      "no_answer": 0,
      "correct_pct": 0.7804878048780488,
      "wrong_pct": 0.024390243902439025,
      "correct_lo": 0.6329469790176472,
      "correct_hi": 0.8799695984461366,
      "wrong_lo": 0.00431842027071648,
      "wrong_hi": 0.12595347011678087,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "closed",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 40,
      "gradeable": 40,
      "correct": 11,
      "incorrect": 0,
      "abstained": 29,
      "no_answer": 0,
      "correct_pct": 0.275,
      "wrong_pct": 0,
      "correct_lo": 0.1610785343314309,
      "correct_hi": 0.4283525083312457,
      "wrong_lo": 0,
      "wrong_hi": 0.08762453925039233,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "mcp",
      "job": "regulatory",
      "kind": "public-verifiable",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 2,
      "questions": 40,
      "gradeable": 80,
      "correct": 35,
      "incorrect": 3,
      "abstained": 42,
      "no_answer": 0,
      "correct_pct": 0.4375,
      "wrong_pct": 0.0375,
      "correct_lo": 0.29602809418250353,
      "correct_hi": 0.5899249732237956,
      "wrong_lo": 0.008707641902894095,
      "wrong_hi": 0.1473450569037188,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "closed",
      "kind": "os-derived",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 1,
      "questions": 13,
      "gradeable": 13,
      "correct": 0,
      "incorrect": 0,
      "abstained": 13,
      "no_answer": 0,
      "correct_pct": 0,
      "wrong_pct": 0,
      "correct_lo": 0,
      "correct_hi": 0.22810184305529166,
      "wrong_lo": 0,
      "wrong_hi": 0.22810184305529166,
      "verdict": "verify"
    },
    {
      "model": "grok-4.7",
      "track": "mcp",
      "kind": "os-derived",
      "provider": "xai",
      "harness": "spacebench-runner",
      "runs": 2,
      "questions": 13,
      "gradeable": 26,
      "correct": 19,
      "incorrect": 2,
      "abstained": 5,
      "no_answer": 0,
      "correct_pct": 0.7307692307692307,
      "wrong_pct": 0.07692307692307693,
      "correct_lo": 0.4598441149709075,
      "correct_hi": 0.896416572849727,
      "wrong_lo": 0.013710066379696295,
      "wrong_hi": 0.3331453392824736,
      "verdict": "verify"
    }
  ]
}