{"abstract":"Only the final request contributes to the batch acceptance total.","category":"Speculative decoding bookkeeping","checks":7,"contract":"For rows of proposed token counts and accepted counts, sum each accepted count after bounding it to that row size.","contract_signature":"proposed, accepted","evaluation_group":"spec-batch-accepted-total","failed_approach":"Summing all proposals charges rejected suffixes.","family":"w3-speculative-decoding-bookkeeping-batch-accepted-total","id":"FA-100706","implementations":{"attempt":{"sha256":"a5b5a229b925bd6923275fb6b79ef6028108de36ab908f9aa76d0b3ee84afb1c","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(proposed, accepted):\n    return sum(proposed)\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('two requests', solve([3,2],[2,1]),3)\ncheck('empty batch', solve([],[]),0)\ncheck('zero accepted', solve([N],[0]),0)\ncheck('oversized row', solve([1,3],[8,2]),3)\ncheck('short counts', solve([2,4],[1]),1)\ncheck('negative count', solve([3],[-1]),0)\ncheck('distinct totals', solve([1,4],[1,3]),4)\nprint(json.dumps({\"observations\": observations, \"passed\": all(x[\"passed\"] for x in observations)}, ensure_ascii=False))\nraise SystemExit(0 if all(x[\"passed\"] for x in observations) else 1)\n"},"broken":{"sha256":"c1b5a7086b8a3971c15d05aeb9c1e35832962d90a32155e6a3685796ff3da5e5","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(proposed, accepted):\n    return accepted[-1] if accepted else 0\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('two requests', solve([3,2],[2,1]),3)\ncheck('empty batch', solve([],[]),0)\ncheck('zero accepted', solve([N],[0]),0)\ncheck('oversized row', solve([1,3],[8,2]),3)\ncheck('short counts', solve([2,4],[1]),1)\ncheck('negative count', solve([3],[-1]),0)\ncheck('distinct totals', solve([1,4],[1,3]),4)\nprint(json.dumps({\"observations\": observations, \"passed\": all(x[\"passed\"] for x in observations)}, ensure_ascii=False))\nraise SystemExit(0 if all(x[\"passed\"] for x in observations) else 1)\n"}},"limitations":"These are finite toy sequences with explicit contracts; they make no claim about production decoder behavior or model quality. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.","method":"Deterministic executable model with adversarial boundary fixtures.","provenance":{"created_by":"Failure Map","dependencies":"Python standard library","family":"w3-speculative-decoding-bookkeeping-batch-accepted-total","generated_at":"2026-09-30T16:48:38.598268+00:00","license":"CC0-1.0","python":"3.12.14","seed":1,"split":"open-access"},"relevance":"A self-contained bookkeeping model makes prefix and token-count semantics observable without a model runtime.","root_cause":"Only the last request contributes to the batch total.","sha256":"db85a7b866d4d3097869c1bd68a18d0d2e9e5606e26e8af3d01e908f787fcf3b","title":"Decoding: Only the final request contributes to the batch acceptance total · case 01","variant":1,"variant_policy":"The fixture code references N, supplied as 1 through 5. Fixed boundary checks may repeat. These records share one mechanism and must remain together in evaluation splits.","verified":true,"visibility":"public","verification":{"attempt":{"elapsed_ms":31.822,"exit_code":1,"observations":[{"actual":5,"check":"two requests","expected":3,"passed":false},{"actual":0,"check":"empty batch","expected":0,"passed":true},{"actual":1,"check":"zero accepted","expected":0,"passed":false},{"actual":4,"check":"oversized row","expected":3,"passed":false},{"actual":6,"check":"short counts","expected":1,"passed":false},{"actual":3,"check":"negative count","expected":0,"passed":false},{"actual":5,"check":"distinct totals","expected":4,"passed":false}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"two requests\", \"actual\": 5, \"expected\": 3, \"passed\": false}, {\"check\": \"empty batch\", \"actual\": 0, \"expected\": 0, \"passed\": true}, {\"check\": \"zero accepted\", \"actual\": 1, \"expected\": 0, \"passed\": false}, {\"check\": \"oversized row\", \"actual\": 4, \"expected\": 3, \"passed\": false}, {\"check\": \"short counts\", \"actual\": 6, \"expected\": 1, \"passed\": false}, {\"check\": \"negative count\", \"actual\": 3, \"expected\": 0, \"passed\": false}, {\"check\": \"distinct totals\", \"actual\": 5, \"expected\": 4, \"passed\": false}], \"passed\": false}\n"},"broken":{"elapsed_ms":31.491,"exit_code":1,"observations":[{"actual":1,"check":"two requests","expected":3,"passed":false},{"actual":0,"check":"empty batch","expected":0,"passed":true},{"actual":0,"check":"zero accepted","expected":0,"passed":true},{"actual":2,"check":"oversized row","expected":3,"passed":false},{"actual":1,"check":"short counts","expected":1,"passed":true},{"actual":-1,"check":"negative count","expected":0,"passed":false},{"actual":3,"check":"distinct totals","expected":4,"passed":false}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"two requests\", \"actual\": 1, \"expected\": 3, \"passed\": false}, {\"check\": \"empty batch\", \"actual\": 0, \"expected\": 0, \"passed\": true}, {\"check\": \"zero accepted\", \"actual\": 0, \"expected\": 0, \"passed\": true}, {\"check\": \"oversized row\", \"actual\": 2, \"expected\": 3, \"passed\": false}, {\"check\": \"short counts\", \"actual\": 1, \"expected\": 1, \"passed\": true}, {\"check\": \"negative count\", \"actual\": -1, \"expected\": 0, \"passed\": false}, {\"check\": \"distinct totals\", \"actual\": 3, \"expected\": 4, \"passed\": false}], \"passed\": false}\n"}},"member_only":{"stages":["fixed"],"fields":["implementations.fixed","verification.fixed","harness","repair"],"note":"The verified repair, its recorded checks, the repair description, and the scoring harness are available to members."}}