{
  "release": "ADO-EVIDENCE-003",
  "title": "Corrected Negative Result in Relational Language Integration",
  "release_date": "2026-07-27",
  "status": "corrected_negative_result",
  "independent_replication": false,
  "original_run": {
    "status": "invalidated_for_interpretation",
    "reason": "The source was played before a memory rung existed, preserving occurrence counts but not source relationships.",
    "summary_sha256": "96941355d52a3153d78611e5af87325fb03e87b4006c7c846486c1e429ce63ac"
  },
  "correction": {
    "online_relational_admission": true,
    "fixed_occurrence_threshold": false,
    "fixed_minimum_vocabulary": false,
    "source_relations_written_during_first_pass": true,
    "source_passes": 1
  },
  "source_measurements": {
    "tokens": 848,
    "unique_and_admitted_tokens": 259,
    "memory_rungs_before_questions": 4,
    "directional_relations_before_questions": 2217,
    "deep_relations_before_questions": 2785,
    "maximum_decision_depth": 6
  },
  "questions_and_outputs": [
    {
      "question": "What does this imply for intelligence and seeking answers?",
      "output": "for a"
    },
    {
      "question": "Pose one question back to probe your ledger dynamics.",
      "output": "the"
    }
  ],
  "preregistered_gates": {
    "integration": false,
    "insight_rather_than_fragment": false,
    "specific_follow_up": false,
    "overall_pass": false
  },
  "interpretation": "The corrected adapter stored relational memory but did not satisfy the frozen language-integration challenge.",
  "commitments": {
    "corrected_summary_sha256": "04fd40ddfa0caca2420ecdda32690f05a077ab52522b4158572f809d53520361",
    "sealed_full_ledger_sha256": "42b59662d40842448d86a7e764d5dbab118d33e7df21ba0244cd0c6101fc3ece",
    "source_vignette_sha256": "a22f9d381df54a234ef3142435987ff401f10951e0dd952f06fc64b36f535b59",
    "amended_protocol_sha256": "a21bbac5b0c85941a037d85081990566f4d2e3fc8383bfd03647aff2c2739253",
    "relational_adapter_sha256": "20f72ee7c914987a528dc19a7ca38b985b4fdbebeb101fadc8688d8857bbf0b2"
  },
  "withheld": [
    "shape, geometry, and topology",
    "protected kernel construction",
    "source code and raw internal traces",
    "equations, mappings, and parameters",
    "reverse-engineering detail"
  ],
  "claim_limits": [
    "This result is not evidence of emergent language learning or generalization.",
    "This result does not establish intelligence, consciousness, preference, or agency.",
    "This result is a corrected negative behavioral result that warrants a better controlled developmental test."
  ]
}