{
 "status": "failed",
 "reasons": [
  "Generated labels disagree with independently re-solved answers.",
  "Unsupported answer for the declared comparator; independent manual verification needed.",
  "Incomplete deterministic re-solve sample."
 ],
 "counts": {
  "generated": 5120,
  "accepted": 640,
  "rejected": 3220,
  "execution_errors": 1260
 },
 "sample_size": 56,
 "label_error_count": 25,
 "overlap": {
  "method": "nfkc_casefold_whitespace_v1",
  "checked_items": 640,
  "evaluation_items": 1319,
  "overlap_count": 0,
  "duplicate_count": 0,
  "matches": []
 },
 "answer_comparison": "numeric_exact",
 "correctness_scope": "Answer agreement with declared independent re-solves; semantic truth is not established."
}