{"abstract":"Two objective buckets share one normalization denominator.","category":"Gradient accumulation weighting","checks":7,"contract":"Each bucket is a separate objective with its own sums and counts. Return its total sum divided by its total count, with zero for an empty bucket.","contract_signature":"bucket_sums, bucket_counts","evaluation_group":"gradient-accumulation-bucketed_objective","failed_approach":"Averaging microbatch means equally retains unequal within-objective batch weighting.","family":"w3-gradient-accumulation-weighting-bucketed-objective","id":"FA-98931","implementations":{"attempt":{"sha256":"6218bf3c08dfe8d2eff83dd014c105475b4b8a904a9b24e31fdeb14ba472a0c2","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(bucket_sums, bucket_counts):\n    return [sum(s/c for s,c in zip(sr,cr) if c)/max(1,sum(c>0 for c in cr)) for sr,cr in zip(bucket_sums,bucket_counts)]\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('independent objectives',solve([[2*N,12*N],[6*N,4*N]],[[N,3*N],[2*N,2*N]]),[3.5,2.5])\ncheck('one item each',solve([[2*N],[8*N]],[[N],[2*N]]),[2,4])\ncheck('empty bucket',solve([[],[6*N]],[[],[2*N]]),[0,3])\ncheck('both empty',solve([[],[]],[[],[]]),[0,0])\ncheck('zero numerator',solve([[0,9*N],[4*N,0]],[[N,3*N],[2*N,2*N]]),[2.25,1])\ncheck('three entries',solve([[N,6*N],[8*N]],[[1,2],[4]]),[7*N/3,2*N])\ncheck('unequal bucket sizes',solve([[4*N],[2*N,8*N]],[[2*N],[N,3*N]]),[2,2.5])\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":"539f54b12c6d073b1029f6bf2167ac93e30445f62ff183492a498993c3c0d435","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(bucket_sums, bucket_counts):\n    denominator=max(1,sum(sum(row) for row in bucket_counts))\n    return [sum(row)/denominator for row in bucket_sums]\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('independent objectives',solve([[2*N,12*N],[6*N,4*N]],[[N,3*N],[2*N,2*N]]),[3.5,2.5])\ncheck('one item each',solve([[2*N],[8*N]],[[N],[2*N]]),[2,4])\ncheck('empty bucket',solve([[],[6*N]],[[],[2*N]]),[0,3])\ncheck('both empty',solve([[],[]],[[],[]]),[0,0])\ncheck('zero numerator',solve([[0,9*N],[4*N,0]],[[N,3*N],[2*N,2*N]]),[2.25,1])\ncheck('three entries',solve([[N,6*N],[8*N]],[[1,2],[4]]),[7*N/3,2*N])\ncheck('unequal bucket sizes',solve([[4*N],[2*N,8*N]],[[2*N],[N,3*N]]),[2,2.5])\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 scalar or short-list summaries and do not model autodiff, devices, or an optimizer. 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-gradient-accumulation-weighting-bucketed-objective","generated_at":"2026-09-30T16:48:18.534175+00:00","license":"CC0-1.0","python":"3.12.14","seed":1,"split":"open-access"},"relevance":"A deterministic toy model isolates one local accumulation contract.","root_cause":"The broken path applies one global count to all objective buckets instead of one denominator per bucket.","sha256":"b42f8092f734400fb4d5e67e396f2927b537bbeb60b3b7c0cb54d671ae3df630","title":"Two objective buckets share one normalization denominator · 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":27.535,"exit_code":1,"observations":[{"actual":[3.0,2.5],"check":"independent objectives","expected":[3.5,2.5],"passed":false},{"actual":[2.0,4.0],"check":"one item each","expected":[2,4],"passed":true},{"actual":[0.0,3.0],"check":"empty bucket","expected":[0,3],"passed":true},{"actual":[0.0,0.0],"check":"both empty","expected":[0,0],"passed":true},{"actual":[1.5,1.0],"check":"zero numerator","expected":[2.25,1],"passed":false},{"actual":[2.0,2.0],"check":"three entries","expected":[2.3333333333333335,2],"passed":false},{"actual":[2.0,2.333333333333333],"check":"unequal bucket sizes","expected":[2,2.5],"passed":false}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"independent objectives\", \"actual\": [3.0, 2.5], \"expected\": [3.5, 2.5], \"passed\": false}, {\"check\": \"one item each\", \"actual\": [2.0, 4.0], \"expected\": [2, 4], \"passed\": true}, {\"check\": \"empty bucket\", \"actual\": [0.0, 3.0], \"expected\": [0, 3], \"passed\": true}, {\"check\": \"both empty\", \"actual\": [0.0, 0.0], \"expected\": [0, 0], \"passed\": true}, {\"check\": \"zero numerator\", \"actual\": [1.5, 1.0], \"expected\": [2.25, 1], \"passed\": false}, {\"check\": \"three entries\", \"actual\": [2.0, 2.0], \"expected\": [2.3333333333333335, 2], \"passed\": false}, {\"check\": \"unequal bucket sizes\", \"actual\": [2.0, 2.333333333333333], \"expected\": [2, 2.5], \"passed\": false}], \"passed\": false}\n"},"broken":{"elapsed_ms":29.912,"exit_code":1,"observations":[{"actual":[1.75,1.25],"check":"independent objectives","expected":[3.5,2.5],"passed":false},{"actual":[0.6666666666666666,2.6666666666666665],"check":"one item each","expected":[2,4],"passed":false},{"actual":[0.0,3.0],"check":"empty bucket","expected":[0,3],"passed":true},{"actual":[0.0,0.0],"check":"both empty","expected":[0,0],"passed":true},{"actual":[1.125,0.5],"check":"zero numerator","expected":[2.25,1],"passed":false},{"actual":[1.0,1.1428571428571428],"check":"three entries","expected":[2.3333333333333335,2],"passed":false},{"actual":[0.6666666666666666,1.6666666666666667],"check":"unequal bucket sizes","expected":[2,2.5],"passed":false}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"independent objectives\", \"actual\": [1.75, 1.25], \"expected\": [3.5, 2.5], \"passed\": false}, {\"check\": \"one item each\", \"actual\": [0.6666666666666666, 2.6666666666666665], \"expected\": [2, 4], \"passed\": false}, {\"check\": \"empty bucket\", \"actual\": [0.0, 3.0], \"expected\": [0, 3], \"passed\": true}, {\"check\": \"both empty\", \"actual\": [0.0, 0.0], \"expected\": [0, 0], \"passed\": true}, {\"check\": \"zero numerator\", \"actual\": [1.125, 0.5], \"expected\": [2.25, 1], \"passed\": false}, {\"check\": \"three entries\", \"actual\": [1.0, 1.1428571428571428], \"expected\": [2.3333333333333335, 2], \"passed\": false}, {\"check\": \"unequal bucket sizes\", \"actual\": [0.6666666666666666, 1.6666666666666667], \"expected\": [2, 2.5], \"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."}}