{"abstract":"The first observed token is paired with itself after start insertion.","category":"Autoregressive target shifting","checks":6,"contract":"Given observations and bos, return adjacent pairs in [bos]+tokens, excluding any pair beyond the final observed target.","contract_signature":"tokens, bos","evaluation_group":"w3-ars-bos_prediction","failed_approach":"Dropping the marker restores observed adjacency but loses the first prediction from context.","family":"w3-autoregressive-target-shifting-bos-context-offset","id":"FA-98256","implementations":{"attempt":{"sha256":"0d15fe16c331adf070b12878f66b32d93db9fe94a8c8982ba9ab13536c0bb006","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(tokens, bos):\n    return [[tokens[i],tokens[i+1]] for i in range(len(tokens)-1)]\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('ordinary',solve([N,N+2],6),[[6,N],[N,N+2]])\ncheck('singleton',solve([4],1),[[1,4]])\ncheck('empty',solve([],1),[])\ncheck('bos equals token',solve([3,4],3),[[3,3],[3,4]])\ncheck('three observations',solve([0,2,5],9),[[9,0],[0,2],[2,5]])\ncheck('large marker',solve([8,7],N+8),[[N+8,8],[8,7]])\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":"046138a7dc8e5dd55f1cb7dabcc4a5474ec4bc520d549419514cba9f4c0a4a67","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(tokens, bos):\n    return [[t,t] for t in tokens]\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('ordinary',solve([N,N+2],6),[[6,N],[N,N+2]])\ncheck('singleton',solve([4],1),[[1,4]])\ncheck('empty',solve([],1),[])\ncheck('bos equals token',solve([3,4],3),[[3,3],[3,4]])\ncheck('three observations',solve([0,2,5],9),[[9,0],[0,2],[2,5]])\ncheck('large marker',solve([8,7],N+8),[[N+8,8],[8,7]])\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":"Inputs are toy integer token sequences; this contract says nothing about model quality, batching, sampling, or production training frameworks. 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-autoregressive-target-shifting-bos-context-offset","generated_at":"2026-09-30T16:48:11.053543+00:00","license":"CC0-1.0","python":"3.12.14","seed":1,"split":"open-access"},"relevance":"A small deterministic token-stream contract makes next-token alignment errors inspectable.","root_cause":"Targets are read from the input position instead of the next position after start-marker insertion.","sha256":"9ccf7fcadb539100894fd36ec02fb694ef8c4cf62b4b2815c95faac3ba906e8c","title":"The first observed token is paired with itself after start insertion · 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":27.51,"exit_code":1,"observations":[{"actual":[[1,3]],"check":"ordinary","expected":[[6,1],[1,3]],"passed":false},{"actual":[],"check":"singleton","expected":[[1,4]],"passed":false},{"actual":[],"check":"empty","expected":[],"passed":true},{"actual":[[3,4]],"check":"bos equals token","expected":[[3,3],[3,4]],"passed":false},{"actual":[[0,2],[2,5]],"check":"three observations","expected":[[9,0],[0,2],[2,5]],"passed":false},{"actual":[[8,7]],"check":"large marker","expected":[[9,8],[8,7]],"passed":false}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"ordinary\", \"actual\": [[1, 3]], \"expected\": [[6, 1], [1, 3]], \"passed\": false}, {\"check\": \"singleton\", \"actual\": [], \"expected\": [[1, 4]], \"passed\": false}, {\"check\": \"empty\", \"actual\": [], \"expected\": [], \"passed\": true}, {\"check\": \"bos equals token\", \"actual\": [[3, 4]], \"expected\": [[3, 3], [3, 4]], \"passed\": false}, {\"check\": \"three observations\", \"actual\": [[0, 2], [2, 5]], \"expected\": [[9, 0], [0, 2], [2, 5]], \"passed\": false}, {\"check\": \"large marker\", \"actual\": [[8, 7]], \"expected\": [[9, 8], [8, 7]], \"passed\": false}], \"passed\": false}\n"},"broken":{"elapsed_ms":25.814,"exit_code":1,"observations":[{"actual":[[1,1],[3,3]],"check":"ordinary","expected":[[6,1],[1,3]],"passed":false},{"actual":[[4,4]],"check":"singleton","expected":[[1,4]],"passed":false},{"actual":[],"check":"empty","expected":[],"passed":true},{"actual":[[3,3],[4,4]],"check":"bos equals token","expected":[[3,3],[3,4]],"passed":false},{"actual":[[0,0],[2,2],[5,5]],"check":"three observations","expected":[[9,0],[0,2],[2,5]],"passed":false},{"actual":[[8,8],[7,7]],"check":"large marker","expected":[[9,8],[8,7]],"passed":false}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"ordinary\", \"actual\": [[1, 1], [3, 3]], \"expected\": [[6, 1], [1, 3]], \"passed\": false}, {\"check\": \"singleton\", \"actual\": [[4, 4]], \"expected\": [[1, 4]], \"passed\": false}, {\"check\": \"empty\", \"actual\": [], \"expected\": [], \"passed\": true}, {\"check\": \"bos equals token\", \"actual\": [[3, 3], [4, 4]], \"expected\": [[3, 3], [3, 4]], \"passed\": false}, {\"check\": \"three observations\", \"actual\": [[0, 0], [2, 2], [5, 5]], \"expected\": [[9, 0], [0, 2], [2, 5]], \"passed\": false}, {\"check\": \"large marker\", \"actual\": [[8, 8], [7, 7]], \"expected\": [[9, 8], [8, 7]], \"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."}}