{
  "schema_version": 2,
  "dataset_version": "1.0.0+5b4307882dc4",
  "release_date": "2026-08-19",
  "purpose": "Hard retrieval benchmark for Pattern Lens using fresh paraphrases, metalinguistic traps and neutral negative cases under the reviewed precision-first reasoning policy.",
  "evidence_limit": "Engineering robustness benchmark. Cases are hand-authored to probe generalisation and abstention; they are not an independent learning-efficacy study or statistical precision/recall evidence.",
  "architecture": "hybrid deterministic retrieval with explicit abstention requirements and categorical reasoning strength",
  "reasoning_strength_policy": "direct > supported > prompt; categorical editorial evidence, not probability",
  "thresholds": {
    "positive_pattern_hit_at_3": 0.9,
    "positive_move_hit_at_3": 0.9,
    "negative_abstention_rate": 0.9,
    "false_positive_rate_max": 0.1
  },
  "metrics": {
    "positive_cases": 25,
    "negative_cases": 15,
    "positive_pattern_hit_at_3": 1,
    "positive_move_hit_at_3": 1,
    "negative_abstention_rate": 0.9333,
    "false_positive_rate": 0.0667,
    "false_positive_cases": [
      "HN14"
    ],
    "positive_misses": []
  },
  "excluded_ambiguous_reasoning_cases": [
    {
      "id": "HP08",
      "language": "en",
      "sentence": "I prefer asynchronous updates to daily meetings.",
      "reason": "preference may retrieve a literal language pattern but is not itself evidence of a completed decision"
    },
    {
      "id": "HP19",
      "language": "de",
      "sentence": "Wenn die Last steigt, würden wir mehr Instanzen brauchen.",
      "reason": "ordinary conditional consequence is not automatically hypothesis testing"
    }
  ],
  "positive_cases": [
    {
      "id": "HP01",
      "language": "en",
      "sentence": "Poor onboarding leads to lower activation.",
      "expected_pattern": "CLF059",
      "expected_move": "Cause",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF059",
          "reasoning_move": "Cause",
          "score": 23.8,
          "evidence_type": "literal+reasoning",
          "hits": [
            "leads to"
          ],
          "reasoning_match": {
            "rule_id": "en-cause-led",
            "intent_id": "connect-cause-and-effect",
            "strength": "direct",
            "evidence": "explicit causal result cue"
          }
        },
        {
          "id": "GRP055",
          "reasoning_move": null,
          "score": 6.2,
          "evidence_type": "literal",
          "hits": [
            "V",
            "V"
          ],
          "reasoning_match": null
        }
      ]
    },
    {
      "id": "HP02",
      "language": "en",
      "sentence": "The launch slipped because legal approval arrived late.",
      "expected_pattern": "CLF060",
      "expected_move": "Cause",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF060",
          "reasoning_move": "Cause",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "en-cause-source",
            "intent_id": "explain-a-cause",
            "strength": "supported",
            "evidence": "explicit causal source"
          }
        },
        {
          "id": "GRP055",
          "reasoning_move": null,
          "score": 6.2,
          "evidence_type": "literal",
          "hits": [
            "V",
            "V"
          ],
          "reasoning_match": null
        }
      ]
    },
    {
      "id": "HP03",
      "language": "de",
      "sentence": "Fehlende Wartung führt zu höheren Ausfallzeiten.",
      "expected_pattern": "CLF059",
      "expected_move": "Cause",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF059",
          "reasoning_move": "Cause",
          "score": 28.3,
          "evidence_type": "literal+reasoning",
          "hits": [
            "führt zu",
            "führt"
          ],
          "reasoning_match": {
            "rule_id": "de-cause-fuehrt",
            "intent_id": "connect-cause-and-effect",
            "strength": "direct",
            "evidence": "führt zu"
          }
        }
      ]
    },
    {
      "id": "HP04",
      "language": "en",
      "sentence": "It seems that the server is overloaded.",
      "expected_pattern": "CLF053",
      "expected_move": "Infer",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF053",
          "reasoning_move": "Infer",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "en-infer-likely",
            "intent_id": "offer-a-likely-explanation",
            "strength": "supported",
            "evidence": "epistemic likelihood cue"
          }
        },
        {
          "id": "GRP055",
          "reasoning_move": null,
          "score": 6.2,
          "evidence_type": "literal",
          "hits": [
            "V",
            "V"
          ],
          "reasoning_match": null
        }
      ]
    },
    {
      "id": "HP05",
      "language": "de",
      "sentence": "Vermutlich liegt die Verzögerung am Netzwerk.",
      "expected_pattern": "CLF053",
      "expected_move": "Infer",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF053",
          "reasoning_move": "Infer",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "de-infer-likely",
            "intent_id": "offer-a-likely-explanation",
            "strength": "supported",
            "evidence": "likelihood cue"
          }
        }
      ]
    },
    {
      "id": "HP06",
      "language": "en",
      "sentence": "Compared with the old process, this one is faster.",
      "expected_pattern": "CLF064",
      "expected_move": "Compare",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF064",
          "reasoning_move": "Compare",
          "score": 24.3,
          "evidence_type": "literal+reasoning",
          "hits": [
            "Compared with"
          ],
          "reasoning_match": {
            "rule_id": "en-compare-compared",
            "intent_id": "compare-alternatives",
            "strength": "direct",
            "evidence": "explicit comparison cue"
          }
        }
      ]
    },
    {
      "id": "HP07",
      "language": "de",
      "sentence": "Der Unterschied zwischen den beiden Ansätzen liegt in den Kosten.",
      "expected_pattern": "CLF063",
      "expected_move": "Compare",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF063",
          "reasoning_move": "Compare",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "de-compare-difference",
            "intent_id": "show-a-difference",
            "strength": "supported",
            "evidence": "explicit difference between alternatives"
          }
        }
      ]
    },
    {
      "id": "HP09",
      "language": "de",
      "sentence": "Wir müssen uns zwischen Geschwindigkeit und Genauigkeit entscheiden.",
      "expected_pattern": "CLF049",
      "expected_move": "Decide",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF049",
          "reasoning_move": "Decide",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "de-decide-between",
            "intent_id": "choose-between-options",
            "strength": "supported",
            "evidence": "explicit choice between alternatives"
          }
        },
        {
          "id": "XPRTRN003",
          "reasoning_move": null,
          "score": 6.6,
          "evidence_type": "literal",
          "hits": [
            "Wir",
            "wir"
          ],
          "reasoning_match": null
        },
        {
          "id": "XPRREG003",
          "reasoning_move": null,
          "score": 3,
          "evidence_type": "literal",
          "hits": [
            "Wir müssen"
          ],
          "reasoning_match": null
        }
      ]
    },
    {
      "id": "HP10",
      "language": "en",
      "sentence": "I see your point, but the assumption still looks risky.",
      "expected_pattern": "CLF058",
      "expected_move": "Challenge",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF058",
          "reasoning_move": "Challenge",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "lens-en-challenge-polite-see",
            "intent_id": "disagree-politely",
            "strength": "supported",
            "evidence": "softened disagreement paraphrase"
          }
        }
      ]
    },
    {
      "id": "HP11",
      "language": "de",
      "sentence": "Ich verstehe deinen Punkt, aber die Daten sprechen dagegen.",
      "expected_pattern": "CLF058",
      "expected_move": "Challenge",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF058",
          "reasoning_move": "Challenge",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "de-challenge-polite",
            "intent_id": "disagree-politely",
            "strength": "supported",
            "evidence": "softened disagreement"
          }
        }
      ]
    },
    {
      "id": "HP12",
      "language": "en",
      "sentence": "It's not a capacity problem, but a coordination problem.",
      "expected_pattern": "CLF061",
      "expected_move": "Reframe",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF061",
          "reasoning_move": "Reframe",
          "score": 18,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "en-reframe-notbut",
            "intent_id": "correct-an-assumption",
            "strength": "direct",
            "evidence": "explicit correction cue"
          }
        }
      ]
    },
    {
      "id": "HP13",
      "language": "de",
      "sentence": "Es geht nicht um mehr Features, sondern um weniger Reibung.",
      "expected_pattern": "CLF061",
      "expected_move": "Reframe",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF061",
          "reasoning_move": "Reframe",
          "score": 18,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "de-reframe-notbut",
            "intent_id": "correct-an-assumption",
            "strength": "direct",
            "evidence": "nicht ... sondern"
          }
        }
      ]
    },
    {
      "id": "HP14",
      "language": "en",
      "sentence": "More documentation alone is not enough.",
      "expected_pattern": "CLF042",
      "expected_move": "Limit",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF042",
          "reasoning_move": "Limit",
          "score": 18,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "en-limit-not-enough",
            "intent_id": "show-something-is-not-enough",
            "strength": "direct",
            "evidence": "insufficiency cue"
          }
        }
      ]
    },
    {
      "id": "HP15",
      "language": "de",
      "sentence": "Mehr Daten allein reicht nicht.",
      "expected_pattern": "CLF042",
      "expected_move": "Limit",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF042",
          "reasoning_move": "Limit",
          "score": 18,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "de-limit-not-enough",
            "intent_id": "show-something-is-not-enough",
            "strength": "direct",
            "evidence": "insufficiency cue"
          }
        }
      ]
    },
    {
      "id": "HP16",
      "language": "en",
      "sentence": "Without a rollback plan, we cannot deploy safely.",
      "expected_pattern": "CLF043",
      "expected_move": "Condition",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF043",
          "reasoning_move": "Condition",
          "score": 28.3,
          "evidence_type": "literal+reasoning",
          "hits": [
            "Without",
            "cannot"
          ],
          "reasoning_match": {
            "rule_id": "en-condition-without",
            "intent_id": "state-a-prerequisite",
            "strength": "direct",
            "evidence": "without plus impossibility"
          }
        }
      ]
    },
    {
      "id": "HP17",
      "language": "de",
      "sentence": "Ohne Freigabe können wir nicht starten.",
      "expected_pattern": "CLF043",
      "expected_move": "Condition",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF043",
          "reasoning_move": "Condition",
          "score": 23.4,
          "evidence_type": "literal+reasoning",
          "hits": [
            "Ohne"
          ],
          "reasoning_match": {
            "rule_id": "lens-de-condition-ohne-koennen",
            "intent_id": "state-a-prerequisite",
            "strength": "direct",
            "evidence": "ohne plus impossibility"
          }
        },
        {
          "id": "XPRTRN003",
          "reasoning_move": null,
          "score": 6.6,
          "evidence_type": "literal",
          "hits": [
            "Wir",
            "wir"
          ],
          "reasoning_match": null
        },
        {
          "id": "QUE012",
          "reasoning_move": null,
          "score": 3,
          "evidence_type": "literal",
          "hits": [
            "Können wir"
          ],
          "reasoning_match": null
        }
      ]
    },
    {
      "id": "HP18",
      "language": "en",
      "sentence": "If demand doubled, would the system still cope?",
      "expected_pattern": "CLF056",
      "expected_move": "Test",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF056",
          "reasoning_move": "Test",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "lens-en-test-conditional-question",
            "intent_id": "test-an-idea",
            "strength": "supported",
            "evidence": "conditional stress-test question"
          }
        }
      ]
    },
    {
      "id": "HP20",
      "language": "en",
      "sentence": "They decided to postpone the migration.",
      "expected_pattern": "CLF069",
      "expected_move": "Decide",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF069",
          "reasoning_move": "Decide",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "en-decide-stated",
            "intent_id": "state-a-decision",
            "strength": "supported",
            "evidence": "explicit completed decision"
          }
        }
      ]
    },
    {
      "id": "HP21",
      "language": "en",
      "sentence": "The difference between the two plans is mainly cost.",
      "expected_pattern": "CLF063",
      "expected_move": "Compare",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF063",
          "reasoning_move": "Compare",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "lens-en-compare-difference",
            "intent_id": "show-a-difference",
            "strength": "supported",
            "evidence": "explicit difference between alternatives"
          }
        }
      ]
    },
    {
      "id": "HP22",
      "language": "en",
      "sentence": "We chose to postpone the launch.",
      "expected_pattern": "CLF069",
      "expected_move": "Decide",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF069",
          "reasoning_move": "Decide",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "lens-en-decide-chose",
            "intent_id": "state-a-decision",
            "strength": "supported",
            "evidence": "completed choice paraphrase"
          }
        }
      ]
    },
    {
      "id": "HP23",
      "language": "en",
      "sentence": "We lack enough evidence to approve this.",
      "expected_pattern": "CLF041",
      "expected_move": "Limit",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF041",
          "reasoning_move": "Limit",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "lens-en-limit-lack-evidence",
            "intent_id": "show-something-is-not-enough",
            "strength": "supported",
            "evidence": "insufficient evidence paraphrase"
          }
        },
        {
          "id": "GRP055",
          "reasoning_move": null,
          "score": 6.2,
          "evidence_type": "literal",
          "hits": [
            "V",
            "V"
          ],
          "reasoning_match": null
        }
      ]
    },
    {
      "id": "HP24",
      "language": "de",
      "sentence": "Es sieht so aus, als ob der Dienst überlastet ist.",
      "expected_pattern": "CLF053",
      "expected_move": "Infer",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF053",
          "reasoning_move": "Infer",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "lens-de-infer-looks",
            "intent_id": "offer-a-likely-explanation",
            "strength": "supported",
            "evidence": "appearance-to-inference paraphrase"
          }
        }
      ]
    },
    {
      "id": "HP25",
      "language": "de",
      "sentence": "Wir haben beschlossen, den Start zu verschieben.",
      "expected_pattern": "CLF069",
      "expected_move": "Decide",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF069",
          "reasoning_move": "Decide",
          "score": 16,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "de-decide-stated",
            "intent_id": "state-a-decision",
            "strength": "supported",
            "evidence": "explicit completed decision"
          }
        },
        {
          "id": "XPRTRN003",
          "reasoning_move": null,
          "score": 6.6,
          "evidence_type": "literal",
          "hits": [
            "Wir",
            "wir"
          ],
          "reasoning_match": null
        }
      ]
    },
    {
      "id": "HP26",
      "language": "en",
      "sentence": "The problem isn't the price; it's the timing.",
      "expected_pattern": "CLF061",
      "expected_move": "Reframe",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF061",
          "reasoning_move": "Reframe",
          "score": 18,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "lens-en-reframe-isnt-its",
            "intent_id": "correct-an-assumption",
            "strength": "direct",
            "evidence": "X-is-not / Y-is correction frame"
          }
        }
      ]
    },
    {
      "id": "HP27",
      "language": "de",
      "sentence": "Das Problem ist nicht der Preis; es ist der Zeitpunkt.",
      "expected_pattern": "CLF061",
      "expected_move": "Reframe",
      "expected_rank": 1,
      "move_rank": 1,
      "top_results": [
        {
          "id": "CLF061",
          "reasoning_move": "Reframe",
          "score": 18,
          "evidence_type": "reasoning",
          "hits": [],
          "reasoning_match": {
            "rule_id": "lens-de-reframe-istnicht-esist",
            "intent_id": "correct-an-assumption",
            "strength": "direct",
            "evidence": "nicht-X / es-ist-Y correction frame"
          }
        },
        {
          "id": "LEX007",
          "reasoning_move": null,
          "score": 3.1,
          "evidence_type": "literal",
          "hits": [
            "das Problem"
          ],
          "reasoning_match": null
        }
      ]
    }
  ],
  "negative_cases": [
    {
      "id": "HN01",
      "language": "en",
      "sentence": "If is a conjunction that introduces a condition.",
      "reason": "metalinguistic mention of a cue",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN02",
      "language": "en",
      "sentence": "The phrase 'only if' appears in the second paragraph.",
      "reason": "quoted cue in a document description",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN03",
      "language": "en",
      "sentence": "The report title is 'Advantages and Disadvantages of Remote Work'.",
      "reason": "decision vocabulary used as a title",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN04",
      "language": "en",
      "sentence": "The word assumption appears twice in this sentence.",
      "reason": "challenge vocabulary discussed as a word",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN05",
      "language": "en",
      "sentence": "We reviewed the impact section yesterday.",
      "reason": "evaluation vocabulary refers to a document section",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN06",
      "language": "en",
      "sentence": "What if is the title of the next chapter.",
      "reason": "what-if cue used as a title",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN07",
      "language": "en",
      "sentence": "Not is an adverb, but is a conjunction.",
      "reason": "not/but tokens described grammatically",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN08",
      "language": "de",
      "sentence": "„Nur wenn“ steht im zweiten Absatz.",
      "reason": "quoted condition cue in document prose",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN09",
      "language": "de",
      "sentence": "Das Wort „Annahme“ ist schwer auszusprechen.",
      "reason": "challenge vocabulary discussed as a word",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN10",
      "language": "de",
      "sentence": "„Es sei denn“ ist ein fester Ausdruck.",
      "reason": "condition cue discussed as an expression",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN11",
      "language": "de",
      "sentence": "Wenn ist eine Konjunktion.",
      "reason": "metalinguistic mention of a cue",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN12",
      "language": "en",
      "sentence": "The team met on Tuesday and discussed the roadmap.",
      "reason": "neutral prose with no target reasoning frame",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN13",
      "language": "de",
      "sentence": "Das Team traf sich am Dienstag und besprach die Roadmap.",
      "reason": "neutral prose with no target reasoning frame",
      "abstained": true,
      "top_results": []
    },
    {
      "id": "HN14",
      "language": "en",
      "sentence": "The server is in the rack next to the router.",
      "reason": "neutral factual prose",
      "abstained": false,
      "top_results": [
        {
          "id": "GRP055",
          "reasoning_move": null,
          "score": 6.2,
          "evidence_type": "literal",
          "hits": [
            "V",
            "V"
          ],
          "reasoning_match": null
        }
      ]
    },
    {
      "id": "HN15",
      "language": "de",
      "sentence": "Der Server steht neben dem Router.",
      "reason": "neutral factual prose",
      "abstained": true,
      "top_results": []
    }
  ]
}