* feat: add cognitive eval harness with CLI integration 20 eval cases across 8 categories (definition, comparison, cause, procedure, recall, correction, verification, unknown). Metrics: intent accuracy, term capture, surface groundedness, versor closure, trace determinism. CLI: `core eval cognition [--json] [--report PATH]`. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat: add operator calibration replay with deterministic grid search Bounded parameter tuning via eval replay evidence. Grid search over salience_top_k and inhibition_threshold with invariant regression guard (versor closure must not regress). Frozen CalibrationParams, before/after metrics, no pack or identity mutation. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
20 lines
5.2 KiB
JSON
20 lines
5.2 KiB
JSON
{"id": "definition_truth_001", "category": "definition", "prompt": "What is truth?", "expected_intent": "definition", "expected_terms": ["truth"], "expected_surface_contains": ["truth"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "definition_light_002", "category": "definition", "prompt": "What is light?", "expected_intent": "definition", "expected_terms": ["light"], "expected_surface_contains": ["light"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "definition_knowledge_003", "category": "definition", "prompt": "What is knowledge?", "expected_intent": "definition", "expected_terms": ["knowledge"], "expected_surface_contains": ["knowledge"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "definition_meaning_004", "category": "definition", "prompt": "What is meaning?", "expected_intent": "definition", "expected_terms": ["meaning"], "expected_surface_contains": ["meaning"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "comparison_truth_light_005", "category": "comparison", "prompt": "Compare truth and light", "expected_intent": "comparison", "expected_terms": ["truth", "light"], "expected_surface_contains": ["truth", "light"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "comparison_word_meaning_006", "category": "comparison", "prompt": "Compare word and meaning", "expected_intent": "comparison", "expected_terms": ["word", "meaning"], "expected_surface_contains": ["word", "meaning"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "cause_light_007", "category": "cause", "prompt": "Why does light exist?", "expected_intent": "cause", "expected_terms": ["light"], "expected_surface_contains": ["light"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "cause_creation_008", "category": "cause", "prompt": "Why does creation matter?", "expected_intent": "cause", "expected_terms": ["creation"], "expected_surface_contains": ["creation"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "cause_wisdom_009", "category": "cause", "prompt": "Why does wisdom matter?", "expected_intent": "cause", "expected_terms": ["wisdom"], "expected_surface_contains": ["wisdom"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "procedure_define_010", "category": "procedure", "prompt": "How do I define a concept?", "expected_intent": "procedure", "expected_terms": ["concept"], "expected_surface_contains": ["concept"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "procedure_compare_011", "category": "procedure", "prompt": "How do I compare two terms?", "expected_intent": "procedure", "expected_terms": ["terms"], "expected_surface_contains": [], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "recall_truth_012", "category": "recall", "prompt": "Remember truth", "expected_intent": "recall", "expected_terms": ["truth"], "expected_surface_contains": ["truth"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "recall_light_013", "category": "recall", "prompt": "Remember light", "expected_intent": "recall", "expected_terms": ["light"], "expected_surface_contains": ["light"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "correction_basic_014", "category": "correction", "prompt": "No, that's wrong", "expected_intent": "correction", "expected_terms": [], "expected_surface_contains": ["correction"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "correction_specific_015", "category": "correction", "prompt": "No, correction means reviewed repair", "expected_intent": "correction", "expected_terms": ["correction"], "expected_surface_contains": ["correction"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "verification_truth_016", "category": "verification", "prompt": "Is truth coherent?", "expected_intent": "verification", "expected_terms": ["truth"], "expected_surface_contains": ["truth"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "verification_light_017", "category": "verification", "prompt": "Is light real?", "expected_intent": "verification", "expected_terms": ["light"], "expected_surface_contains": ["light"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "unknown_word_018", "category": "unknown", "prompt": "word beginning truth", "expected_intent": "unknown", "expected_terms": ["word", "truth"], "expected_surface_contains": [], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "unknown_logos_019", "category": "unknown", "prompt": "light logos", "expected_intent": "unknown", "expected_terms": ["light"], "expected_surface_contains": [], "requires_versor_closure": true, "requires_deterministic_trace": true}
|
|
{"id": "definition_wisdom_020", "category": "definition", "prompt": "What is wisdom?", "expected_intent": "definition", "expected_terms": ["wisdom"], "expected_surface_contains": ["wisdom"], "requires_versor_closure": true, "requires_deterministic_trace": true}
|