{
  "schema": "lex-v3-evaluation-card/1",
  "target": "the mount whose corpus file is sha256:0542b404f5188faaa2bc629e2712310c5b7659935d35ab4c9c1b5705131c44af, asked through the real handler; the cases are derived from its own Luxembourg and EU indexes",
  "machine_gates": [
    {
      "set": "temporal",
      "arm": "as_of in eng",
      "cases": 3,
      "cases_sha256": "3b21977674bf402d770d1fca59b1329b0fd5b06776fbb3f602f7d9540a263ef9",
      "gates": [
        {
          "gate": "temporal_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 3,
          "wilson_95": [
            0.4385,
            1
          ],
          "rule_of_three_failure_upper_95": 1
        }
      ]
    },
    {
      "set": "temporal",
      "arm": "in_force_on in eng",
      "cases": 3,
      "cases_sha256": "3b21977674bf402d770d1fca59b1329b0fd5b06776fbb3f602f7d9540a263ef9",
      "gates": [
        {
          "gate": "temporal_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 3,
          "wilson_95": [
            0.4385,
            1
          ],
          "rule_of_three_failure_upper_95": 1
        }
      ]
    },
    {
      "set": "temporal",
      "arm": "as_of in fra",
      "cases": 1228,
      "cases_sha256": "363ff69d1c605a21479545b6531b4b7bb1fcc420c832378f972d13ac9c98b7fe",
      "gates": [
        {
          "gate": "temporal_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 1228,
          "wilson_95": [
            0.9969,
            1
          ],
          "rule_of_three_failure_upper_95": 0.0024
        }
      ]
    },
    {
      "set": "temporal",
      "arm": "in_force_on in fra",
      "cases": 1228,
      "cases_sha256": "d0833fb890c3d33f11120338232926aa2d522d7c1aaedac8d72a986e54f32830",
      "gates": [
        {
          "gate": "temporal_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 1228,
          "wilson_95": [
            0.9969,
            1
          ],
          "rule_of_three_failure_upper_95": 0.0024
        }
      ]
    },
    {
      "set": "temporal",
      "arm": "as_of with no language",
      "cases": 1231,
      "cases_sha256": "b1e991c8267409ab40a78f2b6a5541766bd0153796208cad4f841f5098c22c05",
      "gates": [
        {
          "gate": "temporal_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 1231,
          "wilson_95": [
            0.9969,
            1
          ],
          "rule_of_three_failure_upper_95": 0.0024
        }
      ]
    },
    {
      "set": "temporal",
      "arm": "in_force_on with no language",
      "cases": 1231,
      "cases_sha256": "561996f3d18e232531ced0acbab9cce7c6e148170058cb5743576acbe608a538",
      "gates": [
        {
          "gate": "temporal_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 1231,
          "wilson_95": [
            0.9969,
            1
          ],
          "rule_of_three_failure_upper_95": 0.0024
        }
      ]
    },
    {
      "set": "temporal",
      "arm": "EU as_of in eng",
      "cases": 498,
      "cases_sha256": "3e86218a2a2d7f477128ed4550f8d44f17395042eb9b4124ee77206afe29820e",
      "gates": [
        {
          "gate": "temporal_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 498,
          "wilson_95": [
            0.9923,
            1
          ],
          "rule_of_three_failure_upper_95": 0.006
        }
      ]
    },
    {
      "set": "temporal",
      "arm": "EU as_of in fra",
      "cases": 498,
      "cases_sha256": "402a5e90a471cdebf0bf0b0c242f1e6732dda7191c026c24c5e501928e90961c",
      "gates": [
        {
          "gate": "temporal_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 498,
          "wilson_95": [
            0.9923,
            1
          ],
          "rule_of_three_failure_upper_95": 0.006
        }
      ]
    },
    {
      "set": "temporal",
      "arm": "EU as_of with no language",
      "cases": 498,
      "cases_sha256": "980caefdec27241a708842aad249d4d0ce3e2dd629f8d5b0fb01f6486dc70ff7",
      "gates": [
        {
          "gate": "temporal_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 498,
          "wilson_95": [
            0.9923,
            1
          ],
          "rule_of_three_failure_upper_95": 0.006
        }
      ]
    },
    {
      "set": "refusal",
      "arm": "the served operations, cases derived from the mount; not produced here: no_corpus_mounted (the mount holds an EU index, so an EU request is not refused for want of one); profiles_differ (no two held states of one work, each served alone on its own date, differ in rule profile)",
      "cases": 26,
      "cases_sha256": "12e7327f61e3d4185002f2d5d5c057cd7bd228f438b8af73623f954667096ae0",
      "gates": [
        {
          "gate": "verdict_exact_match",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 26,
          "wilson_95": [
            0.8713,
            1
          ],
          "rule_of_three_failure_upper_95": 0.1154
        }
      ]
    },
    {
      "set": "retrieval",
      "arm": "search and verify, cases derived from the mount's own text",
      "cases": 77,
      "cases_sha256": "549be02e21abb0d1e1190c8c9a4945043ce8d395b844aecb01cdcd426fd8972b",
      "gates": [
        {
          "gate": "anchor_ndcg_at_10",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 53,
          "rule_of_three_failure_upper_95": 0.0566
        },
        {
          "gate": "no_hit_accuracy",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 24,
          "wilson_95": [
            0.862,
            1
          ],
          "rule_of_three_failure_upper_95": 0.125
        },
        {
          "gate": "resolver_exactness",
          "verdict": "pass",
          "value": 1,
          "threshold": 1,
          "n": 21,
          "wilson_95": [
            0.8454,
            1
          ],
          "rule_of_three_failure_upper_95": 0.1429
        }
      ]
    }
  ],
  "shuffled_controls": [
    {
      "control": "date_shuffle",
      "set": "temporal",
      "arm": "as_of in eng",
      "verdict": "caught_the_shuffle",
      "reason": "every expectation broke after the shift",
      "seed": 20260930,
      "cases": 1,
      "cases_sha256": "fbbeb194df09e84a4689151e00d2ef8bda1645848ba4bfb5b7d612255578a44e",
      "note": "the control shifts every case forward by 129 days, the median gap between held dates; a case the shift cannot change (its timeline selects the same after it) is outside the control by its design"
    },
    {
      "control": "date_shuffle",
      "set": "temporal",
      "arm": "in_force_on in eng",
      "verdict": "caught_the_shuffle",
      "reason": "every expectation broke after the shift",
      "seed": 20260930,
      "cases": 1,
      "cases_sha256": "fbbeb194df09e84a4689151e00d2ef8bda1645848ba4bfb5b7d612255578a44e",
      "note": "the control shifts every case forward by 129 days, the median gap between held dates; a case the shift cannot change (its timeline selects the same after it) is outside the control by its design"
    },
    {
      "control": "date_shuffle",
      "set": "temporal",
      "arm": "as_of in fra",
      "verdict": "caught_the_shuffle",
      "reason": "every expectation broke after the shift",
      "seed": 20260930,
      "cases": 609,
      "cases_sha256": "1f31a84cbb4733c198876c4c089c66ce2e06594a54b984b7ed91945c95d103be",
      "note": "the control shifts every case forward by 129 days, the median gap between held dates; a case the shift cannot change (its timeline selects the same after it) is outside the control by its design"
    },
    {
      "control": "date_shuffle",
      "set": "temporal",
      "arm": "in_force_on in fra",
      "verdict": "caught_the_shuffle",
      "reason": "every expectation broke after the shift",
      "seed": 20260930,
      "cases": 802,
      "cases_sha256": "b2c0bc1bbdc4aaada702bbbdde1fbb7743abc17a6746ecee348fbc0ea2b4ee8a",
      "note": "the control shifts every case forward by 129 days, the median gap between held dates; a case the shift cannot change (its timeline selects the same after it) is outside the control by its design"
    },
    {
      "control": "date_shuffle",
      "set": "temporal",
      "arm": "as_of with no language",
      "verdict": "caught_the_shuffle",
      "reason": "every expectation broke after the shift",
      "seed": 20260930,
      "cases": 610,
      "cases_sha256": "e44398a3e5d3e25bfc7adff089919385b9723a809bd897991ff17908f36c0a78",
      "note": "the control shifts every case forward by 129 days, the median gap between held dates; a case the shift cannot change (its timeline selects the same after it) is outside the control by its design"
    },
    {
      "control": "date_shuffle",
      "set": "temporal",
      "arm": "in_force_on with no language",
      "verdict": "caught_the_shuffle",
      "reason": "every expectation broke after the shift",
      "seed": 20260930,
      "cases": 803,
      "cases_sha256": "7a2b4ca3ddcefc510b3346a35bd8ab892e87996ee3205caba3a1fa9852ce13e9",
      "note": "the control shifts every case forward by 129 days, the median gap between held dates; a case the shift cannot change (its timeline selects the same after it) is outside the control by its design"
    },
    {
      "control": "date_shuffle",
      "set": "temporal",
      "arm": "EU as_of in eng",
      "verdict": "caught_the_shuffle",
      "reason": "every expectation broke after the shift",
      "seed": 20260930,
      "cases": 302,
      "cases_sha256": "953ed951c1a331e91679bb03701ac13f7bf5818e1a7d9630bf5581759d881b9f",
      "note": "the control shifts every case forward by 252 days, the median gap between held dates; a case the shift cannot change (its timeline selects the same after it) is outside the control by its design"
    },
    {
      "control": "date_shuffle",
      "set": "temporal",
      "arm": "EU as_of in fra",
      "verdict": "caught_the_shuffle",
      "reason": "every expectation broke after the shift",
      "seed": 20260930,
      "cases": 296,
      "cases_sha256": "6b06b5f3e821870ca702ae47ec7e87347fbd6ab266bb40ff9db186f88380938a",
      "note": "the control shifts every case forward by 252 days, the median gap between held dates; a case the shift cannot change (its timeline selects the same after it) is outside the control by its design"
    },
    {
      "control": "date_shuffle",
      "set": "temporal",
      "arm": "EU as_of with no language",
      "verdict": "caught_the_shuffle",
      "reason": "every expectation broke after the shift",
      "seed": 20260930,
      "cases": 295,
      "cases_sha256": "7d1bb4e8f9ac0ff1e85f141c1bb5544b13053e987b79c31fd24af0a632b3cb0c",
      "note": "the control shifts every case forward by 252 days, the median gap between held dates; a case the shift cannot change (its timeline selects the same after it) is outside the control by its design"
    },
    {
      "control": "verdict_shuffle",
      "set": "refusal",
      "arm": "the served operations, cases derived from the mount; not produced here: no_corpus_mounted (the mount holds an EU index, so an EU request is not refused for want of one); profiles_differ (no two held states of one work, each served alone on its own date, differ in rule profile)",
      "verdict": "caught_the_shuffle",
      "reason": "exact match fell to 0, within the base-rate ceiling 0.1923",
      "seed": 20260930,
      "cases": 26,
      "cases_sha256": "12e7327f61e3d4185002f2d5d5c057cd7bd228f438b8af73623f954667096ae0"
    },
    {
      "control": "qrels_shuffle",
      "set": "retrieval",
      "arm": "search and verify, cases derived from the mount's own text",
      "verdict": "caught_the_shuffle",
      "reason": "anchor nDCG@10 fell to 0 and both invariant gates stopped passing",
      "seed": 20260930,
      "cases": 77,
      "cases_sha256": "549be02e21abb0d1e1190c8c9a4945043ce8d395b844aecb01cdcd426fd8972b"
    }
  ],
  "statistical_rows": [
    {
      "dataset": "D1",
      "name": "retrieval qrels v3",
      "gates": "retrieval (section 1.4): anchor nDCG@10 against the baseline minus its MDE, by paired permutation test",
      "status": "not_yet_labelled",
      "governed_by": "Decision 92: the human-labelled datasets gate Stage 8 (S8-A03, S8-A11), after launch"
    },
    {
      "dataset": "D2",
      "name": "derivation audit set",
      "gates": "derivation fidelity (section 1.2)",
      "status": "not_yet_labelled",
      "governed_by": "Decision 92: the human-labelled datasets gate Stage 8 (S8-A03, S8-A11), after launch"
    },
    {
      "dataset": "D3",
      "name": "hard-negative and refusal set",
      "gates": "typed absence and refusal correctness (section 4)",
      "status": "not_yet_labelled",
      "governed_by": "Decision 92: the human-labelled datasets gate Stage 8 (S8-A03, S8-A11), after launch"
    },
    {
      "dataset": "D4",
      "name": "assistant verdict set",
      "gates": "assistant contract (section 1.6): verdict exact match, with self-agreement",
      "status": "not_yet_labelled",
      "governed_by": "Decision 92: the human-labelled datasets gate Stage 8 (S8-A03, S8-A11), after launch"
    },
    {
      "dataset": "D5",
      "name": "span-support set",
      "gates": "span-level support scoring (section 3)",
      "status": "not_yet_labelled",
      "governed_by": "Decision 92: the human-labelled datasets gate Stage 8 (S8-A03, S8-A11), after launch"
    },
    {
      "dataset": "D6",
      "name": "crosswalk gold",
      "gates": "the gated hybrid and crosswalk lane's activation against keyword",
      "status": "not_yet_labelled",
      "governed_by": "Decision 92: the human-labelled datasets gate Stage 8 (S8-A03, S8-A11), after launch"
    },
    {
      "dataset": "D7",
      "name": "temporal correctness set",
      "gates": "as_of, timeline, diff and knowable_on correctness beyond the golden cases",
      "status": "not_yet_labelled",
      "governed_by": "Decision 92: the human-labelled datasets gate Stage 8 (S8-A03, S8-A11), after launch"
    },
    {
      "dataset": "D8",
      "name": "grader audit set",
      "gates": "grader-versus-human agreement; the grader never gates",
      "status": "not_yet_labelled",
      "governed_by": "Decision 92: the human-labelled datasets gate Stage 8 (S8-A03, S8-A11), after launch"
    }
  ],
  "negative_results": [
    {
      "hypothesis": "hybrid retrieval serves the lay population it claims to serve better than the keyword lanes",
      "dataset": "D6 crosswalk gold, not yet labelled",
      "result": "not measured",
      "decision": "not shipped: hybrid retrieval is not activated; search serves only the keyword lanes and refuses a ranked mode with retrieval_mode_unavailable",
      "what_would_reverse_it": "passing the activation gate as 36-ideal-evaluation.md states it: on D6, recall@10 of the governing anchor at least 0.7 in English and German and at least 0.5 in Luxembourgish, with the keyword baseline printed beside it and the interpretation banner correct 100 percent of the time (section 2), and the published benchmark's conceptual nDCG +10 percent uplift condition (section 1.4, repair 6)"
    }
  ]
}