{"abstract":"Dot product omits input zero-point correction.","category":"Model quantization arithmetic","checks":6,"contract":"For equal-length code vectors compute sum((a_i-zero_a)*(b_i-zero_b)); inputs are small toy integers. Inputs follow this explicit integer toy contract; ties use Python round-to-even where rounding is required.","contract_signature":"a, b, zero_a, zero_b","evaluation_group":"quant-dot_centering","failed_approach":"The attempted code centers the left operand only; right-side zero point still contributes a spurious term.","family":"w3-model-quantization-arithmetic-dot-zero-correction","id":"FA-98716","implementations":{"attempt":{"sha256":"5a6a0cbbf45aba37888008867a8520b29be07a7ff591205379abff84a274ec07","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(a, b, zero_a, zero_b):\n    return sum((x-zero_a)*y for x,y in zip(a,b))\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('both offsets',solve([N+2, N],[N+1,N+3],N,N),2)\ncheck('zero offsets',solve([1,2],[3,4],0,0),11)\ncheck('empty vectors',solve([],[],N,N),0)\ncheck('zero centered',solve([N],[N],N,N),0)\ncheck('negative centered',solve([N-1],[N+2],N,N),-2)\ncheck('two terms',solve([N+1,N+2],[N+3,N+4],N,N),11)\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":"d8117875dab80398f7c6160fd7c9ad4c7ef920f5237175db755765ba0ddb38ab","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(a, b, zero_a, zero_b):\n    return sum(x*y for x,y in zip(a,b))\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('both offsets',solve([N+2, N],[N+1,N+3],N,N),2)\ncheck('zero offsets',solve([1,2],[3,4],0,0),11)\ncheck('empty vectors',solve([],[],N,N),0)\ncheck('zero centered',solve([N],[N],N,N),0)\ncheck('negative centered',solve([N-1],[N+2],N,N),-2)\ncheck('two terms',solve([N+1,N+2],[N+3,N+4],N,N),11)\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":"Toy scalar/list arithmetic only; it does not model a production quantizer, hardware instruction, or accuracy/performance gain. 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-model-quantization-arithmetic-dot-zero-correction","generated_at":"2026-09-30T16:48:15.842493+00:00","license":"CC0-1.0","python":"3.12.14","seed":1,"split":"open-access"},"relevance":"This deterministic model isolates one quantization arithmetic contract from tensor storage and runtime kernels.","root_cause":"For equal-length code vectors compute sum((a_i-zero_a)*(b_i-zero_b)); inputs are small toy integers.","sha256":"8ff5aa270b56d8a26fb872c4528e5e17f773b80e17fa6583029790013b2ab48f","title":"Quantization: Dot product omits input zero-point correction · 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":30.464,"exit_code":1,"observations":[{"actual":4,"check":"both offsets","expected":2,"passed":false},{"actual":11,"check":"zero offsets","expected":11,"passed":true},{"actual":0,"check":"empty vectors","expected":0,"passed":true},{"actual":0,"check":"zero centered","expected":0,"passed":true},{"actual":-3,"check":"negative centered","expected":-2,"passed":false},{"actual":14,"check":"two terms","expected":11,"passed":false}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"both offsets\", \"actual\": 4, \"expected\": 2, \"passed\": false}, {\"check\": \"zero offsets\", \"actual\": 11, \"expected\": 11, \"passed\": true}, {\"check\": \"empty vectors\", \"actual\": 0, \"expected\": 0, \"passed\": true}, {\"check\": \"zero centered\", \"actual\": 0, \"expected\": 0, \"passed\": true}, {\"check\": \"negative centered\", \"actual\": -3, \"expected\": -2, \"passed\": false}, {\"check\": \"two terms\", \"actual\": 14, \"expected\": 11, \"passed\": false}], \"passed\": false}\n"},"broken":{"elapsed_ms":26.626,"exit_code":1,"observations":[{"actual":10,"check":"both offsets","expected":2,"passed":false},{"actual":11,"check":"zero offsets","expected":11,"passed":true},{"actual":0,"check":"empty vectors","expected":0,"passed":true},{"actual":1,"check":"zero centered","expected":0,"passed":false},{"actual":0,"check":"negative centered","expected":-2,"passed":false},{"actual":23,"check":"two terms","expected":11,"passed":false}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"both offsets\", \"actual\": 10, \"expected\": 2, \"passed\": false}, {\"check\": \"zero offsets\", \"actual\": 11, \"expected\": 11, \"passed\": true}, {\"check\": \"empty vectors\", \"actual\": 0, \"expected\": 0, \"passed\": true}, {\"check\": \"zero centered\", \"actual\": 1, \"expected\": 0, \"passed\": false}, {\"check\": \"negative centered\", \"actual\": 0, \"expected\": -2, \"passed\": false}, {\"check\": \"two terms\", \"actual\": 23, \"expected\": 11, \"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."}}