{"abstract":"Very high earners with children report a negative credit.","category":"Tax bracket computation","checks":8,"contract":"solve(kids, others, magi, status, tax, earned): stipulated child credit, whole dollars. Credit = 2000 per child + 500 per other dependent, reduced by 50 for each 1000 or fraction thereof of MAGI above the threshold (single 200000, mfj 400000), floored at 0. The credit first offsets tax (used = min(credit, tax)); the refundable portion is min(credit - used, 1700 per child, 15% of earned income above 2500 rounded down, floored at 0). Return [credit, used, refundable].","contract_signature":"kids, others, magi, status, tax, earned","evaluation_group":"w2-tax_bracket_computation-child-credit-phaseout","failed_approach":"Exempting the other-dependent part from the phase-out is not the stipulated rule and still misstates the credit.","family":"w2-tax_bracket_computation-child-credit-phaseout-phaseout-floor","id":"FA-62826","implementations":{"attempt":{"sha256":"5a67fd1fd5bd2efb86ab5b7a6449947fb835657a6be5aee3642297d54e93a21d","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(kids, others, magi, status, tax, earned):\n    thr = {'single': 200000, 'mfj': 400000}[status]\n    credit = 2000 * kids + 500 * others\n    if magi > thr: credit = max(500 * others, credit - 50 * (-(-(magi - thr) // 1000)))\n    used = min(credit, tax)\n    refundable = min(credit - used, 1700 * kids, max(0, (earned - 2500) * 15 // 100))\n    return [credit, used, refundable]\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncases = [[('regression phaseout-floor 1', (1, 0, 450000, 'mfj', 5000, 450000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (0, 0, 450000, 'single', 500, 17366), [0, 0, 0]),\n  ('partial repair guard 1', (1, 2, 450000, 'mfj', 5000, 450000), [500, 500, 0]),\n  ('partial repair guard 2', (0, 2, 400001, 'mfj', 500, 3000), [950, 500, 0]),\n  ('control: refund limited by child cap', (2, 0, 150000, 'single', 1000, 30000), [4000, 1000, 3000]),\n  ('control: fraction thereof', (2, 1, 201001, 'single', 10000, 200000), [4400, 4400, 0]),\n  ('control: earned limit', (1, 0, 30000, 'mfj', 0, 10000), [2000, 0, 1125]),\n  ('control: joint partial phase-out', (3, 2, 410500, 'mfj', 1000, 410500), [6450, 1000, 5100])],\n [('regression phaseout-floor 1', (1, 0, 400001, 'single', 0, 10000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (2, 0, 400001, 'single', 500, 0), [0, 0, 0]),\n  ('partial repair guard 1', (3, 2, 400001, 'single', 5000, 10000), [0, 0, 0]),\n  ('partial repair guard 2', (2, 2, 450000, 'single', 5000, 47904), [0, 0, 0]),\n  ('control: fully phased', (1, 0, 450000, 'mfj', 5000, 450000), [0, 0, 0]),\n  ('control: joint partial phase-out', (3, 2, 410500, 'mfj', 1000, 410500), [6450, 1000, 5100]),\n  ('control: other dependents only', (0, 3, 50000, 'single', 0, 50000), [1500, 0, 0]),\n  ('control: low earner', (2, 0, 2000, 'single', 0, 2000), [4000, 0, 0])],\n [('regression phaseout-floor 1', (3, 2, 400001, 'single', 5000, 10000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (2, 2, 450000, 'single', 5000, 47904), [0, 0, 0]),\n  ('partial repair guard 1', (3, 2, 400001, 'single', 1000, 30000), [0, 0, 0]),\n  ('partial repair guard 2', (2, 2, 400001, 'single', 0, 2000), [0, 0, 0]),\n  ('control: low earner', (2, 0, 2000, 'single', 0, 2000), [4000, 0, 0]),\n  ('control: refund limited by child cap', (2, 0, 150000, 'single', 1000, 30000), [4000, 1000, 3000]),\n  ('control: fraction thereof', (2, 1, 201001, 'single', 10000, 200000), [4400, 4400, 0]),\n  ('control: earned limit', (1, 0, 30000, 'mfj', 0, 10000), [2000, 0, 1125])],\n [('regression phaseout-floor 1', (2, 0, 400001, 'single', 10000, 30000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (1, 0, 450000, 'mfj', 10000, 63223), [0, 0, 0]),\n  ('partial repair guard 1', (0, 1, 266447, 'single', 1000, 0), [0, 0, 0]),\n  ('partial repair guard 2', (1, 1, 450000, 'mfj', 0, 10000), [0, 0, 0]),\n  ('control: earned limit', (1, 0, 30000, 'mfj', 0, 10000), [2000, 0, 1125]),\n  ('control: fully phased', (1, 0, 450000, 'mfj', 5000, 450000), [0, 0, 0]),\n  ('control: joint partial phase-out', (3, 2, 410500, 'mfj', 1000, 410500), [6450, 1000, 5100]),\n  ('control: other dependents only', (0, 3, 50000, 'single', 0, 50000), [1500, 0, 0])],\n [('regression phaseout-floor 1', (3, 2, 400001, 'single', 1000, 30000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (2, 2, 400001, 'single', 0, 2000), [0, 0, 0]),\n  ('partial repair guard 1', (1, 2, 450000, 'mfj', 0, 10000), [500, 0, 500]),\n  ('partial repair guard 2', (0, 1, 450000, 'mfj', 5000, 85969), [0, 0, 0]),\n  ('control: other dependents only', (0, 3, 50000, 'single', 0, 50000), [1500, 0, 0]),\n  ('control: low earner', (2, 0, 2000, 'single', 0, 2000), [4000, 0, 0]),\n  ('control: refund limited by child cap', (2, 0, 150000, 'single', 1000, 30000), [4000, 1000, 3000]),\n  ('control: fraction thereof', (2, 1, 201001, 'single', 10000, 200000), [4400, 4400, 0])]]\nfor label, args, expected in cases[N - 1]:\n    check(label, solve(*args), expected)\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":"c69754cf70679a50b03e9246a6d12c5e835988b6696f62bbc765d5427581a57f","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(kids, others, magi, status, tax, earned):\n    thr = {'single': 200000, 'mfj': 400000}[status]\n    credit = 2000 * kids + 500 * others\n    if magi > thr: credit = credit - 50 * (-(-(magi - thr) // 1000))\n    used = min(credit, tax)\n    refundable = min(credit - used, 1700 * kids, max(0, (earned - 2500) * 15 // 100))\n    return [credit, used, refundable]\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncases = [[('regression phaseout-floor 1', (1, 0, 450000, 'mfj', 5000, 450000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (0, 0, 450000, 'single', 500, 17366), [0, 0, 0]),\n  ('partial repair guard 1', (1, 2, 450000, 'mfj', 5000, 450000), [500, 500, 0]),\n  ('partial repair guard 2', (0, 2, 400001, 'mfj', 500, 3000), [950, 500, 0]),\n  ('control: refund limited by child cap', (2, 0, 150000, 'single', 1000, 30000), [4000, 1000, 3000]),\n  ('control: fraction thereof', (2, 1, 201001, 'single', 10000, 200000), [4400, 4400, 0]),\n  ('control: earned limit', (1, 0, 30000, 'mfj', 0, 10000), [2000, 0, 1125]),\n  ('control: joint partial phase-out', (3, 2, 410500, 'mfj', 1000, 410500), [6450, 1000, 5100])],\n [('regression phaseout-floor 1', (1, 0, 400001, 'single', 0, 10000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (2, 0, 400001, 'single', 500, 0), [0, 0, 0]),\n  ('partial repair guard 1', (3, 2, 400001, 'single', 5000, 10000), [0, 0, 0]),\n  ('partial repair guard 2', (2, 2, 450000, 'single', 5000, 47904), [0, 0, 0]),\n  ('control: fully phased', (1, 0, 450000, 'mfj', 5000, 450000), [0, 0, 0]),\n  ('control: joint partial phase-out', (3, 2, 410500, 'mfj', 1000, 410500), [6450, 1000, 5100]),\n  ('control: other dependents only', (0, 3, 50000, 'single', 0, 50000), [1500, 0, 0]),\n  ('control: low earner', (2, 0, 2000, 'single', 0, 2000), [4000, 0, 0])],\n [('regression phaseout-floor 1', (3, 2, 400001, 'single', 5000, 10000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (2, 2, 450000, 'single', 5000, 47904), [0, 0, 0]),\n  ('partial repair guard 1', (3, 2, 400001, 'single', 1000, 30000), [0, 0, 0]),\n  ('partial repair guard 2', (2, 2, 400001, 'single', 0, 2000), [0, 0, 0]),\n  ('control: low earner', (2, 0, 2000, 'single', 0, 2000), [4000, 0, 0]),\n  ('control: refund limited by child cap', (2, 0, 150000, 'single', 1000, 30000), [4000, 1000, 3000]),\n  ('control: fraction thereof', (2, 1, 201001, 'single', 10000, 200000), [4400, 4400, 0]),\n  ('control: earned limit', (1, 0, 30000, 'mfj', 0, 10000), [2000, 0, 1125])],\n [('regression phaseout-floor 1', (2, 0, 400001, 'single', 10000, 30000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (1, 0, 450000, 'mfj', 10000, 63223), [0, 0, 0]),\n  ('partial repair guard 1', (0, 1, 266447, 'single', 1000, 0), [0, 0, 0]),\n  ('partial repair guard 2', (1, 1, 450000, 'mfj', 0, 10000), [0, 0, 0]),\n  ('control: earned limit', (1, 0, 30000, 'mfj', 0, 10000), [2000, 0, 1125]),\n  ('control: fully phased', (1, 0, 450000, 'mfj', 5000, 450000), [0, 0, 0]),\n  ('control: joint partial phase-out', (3, 2, 410500, 'mfj', 1000, 410500), [6450, 1000, 5100]),\n  ('control: other dependents only', (0, 3, 50000, 'single', 0, 50000), [1500, 0, 0])],\n [('regression phaseout-floor 1', (3, 2, 400001, 'single', 1000, 30000), [0, 0, 0]),\n  ('regression phaseout-floor 2', (2, 2, 400001, 'single', 0, 2000), [0, 0, 0]),\n  ('partial repair guard 1', (1, 2, 450000, 'mfj', 0, 10000), [500, 0, 500]),\n  ('partial repair guard 2', (0, 1, 450000, 'mfj', 5000, 85969), [0, 0, 0]),\n  ('control: other dependents only', (0, 3, 50000, 'single', 0, 50000), [1500, 0, 0]),\n  ('control: low earner', (2, 0, 2000, 'single', 0, 2000), [4000, 0, 0]),\n  ('control: refund limited by child cap', (2, 0, 150000, 'single', 1000, 30000), [4000, 1000, 3000]),\n  ('control: fraction thereof', (2, 1, 201001, 'single', 10000, 200000), [4400, 4400, 0])]]\nfor label, args, expected in cases[N - 1]:\n    check(label, solve(*args), expected)\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":"A deterministic, bounded teaching model with a stipulated toy contract; it makes no claim of conformance to any real regulation, standard, or institution's rules. 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":"w2-tax_bracket_computation-child-credit-phaseout-phaseout-floor","generated_at":"2026-09-29T14:47:08.346885+00:00","license":"CC0-1.0","python":"3.12.14","seed":1,"split":"open-access"},"relevance":"Tax computations hinge on which slice, threshold, ordering and rounding rule applies at each step; a misplaced boundary silently misstates liabilities.","root_cause":"The reduced credit is not floored at zero.","sha256":"b09e6ae1ab6c1bc5bd7f5a89e36a2b71ee096b2f99724229eda3fab6269b2a75","title":"The phase-out drives the credit negative · case 01","variant":1,"variant_policy":"Five numbered records share a model and may reuse boundary fixtures.","verified":true,"visibility":"public","verification":{"attempt":{"elapsed_ms":41.762,"exit_code":1,"observations":[{"actual":[0,0,0],"check":"regression phaseout-floor 1","expected":[0,0,0],"passed":true},{"actual":[0,0,0],"check":"regression phaseout-floor 2","expected":[0,0,0],"passed":true},{"actual":[1000,1000,0],"check":"partial repair guard 1","expected":[500,500,0],"passed":false},{"actual":[1000,500,0],"check":"partial repair guard 2","expected":[950,500,0],"passed":false},{"actual":[4000,1000,3000],"check":"control: refund limited by child cap","expected":[4000,1000,3000],"passed":true},{"actual":[4400,4400,0],"check":"control: fraction thereof","expected":[4400,4400,0],"passed":true},{"actual":[2000,0,1125],"check":"control: earned limit","expected":[2000,0,1125],"passed":true},{"actual":[6450,1000,5100],"check":"control: joint partial phase-out","expected":[6450,1000,5100],"passed":true}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"regression phaseout-floor 1\", \"actual\": [0, 0, 0], \"expected\": [0, 0, 0], \"passed\": true}, {\"check\": \"regression phaseout-floor 2\", \"actual\": [0, 0, 0], \"expected\": [0, 0, 0], \"passed\": true}, {\"check\": \"partial repair guard 1\", \"actual\": [1000, 1000, 0], \"expected\": [500, 500, 0], \"passed\": false}, {\"check\": \"partial repair guard 2\", \"actual\": [1000, 500, 0], \"expected\": [950, 500, 0], \"passed\": false}, {\"check\": \"control: refund limited by child cap\", \"actual\": [4000, 1000, 3000], \"expected\": [4000, 1000, 3000], \"passed\": true}, {\"check\": \"control: fraction thereof\", \"actual\": [4400, 4400, 0], \"expected\": [4400, 4400, 0], \"passed\": true}, {\"check\": \"control: earned limit\", \"actual\": [2000, 0, 1125], \"expected\": [2000, 0, 1125], \"passed\": true}, {\"check\": \"control: joint partial phase-out\", \"actual\": [6450, 1000, 5100], \"expected\": [6450, 1000, 5100], \"passed\": true}], \"passed\": false}\n"},"broken":{"elapsed_ms":43.728,"exit_code":1,"observations":[{"actual":[-500,-500,0],"check":"regression phaseout-floor 1","expected":[0,0,0],"passed":false},{"actual":[-12500,-12500,0],"check":"regression phaseout-floor 2","expected":[0,0,0],"passed":false},{"actual":[500,500,0],"check":"partial repair guard 1","expected":[500,500,0],"passed":true},{"actual":[950,500,0],"check":"partial repair guard 2","expected":[950,500,0],"passed":true},{"actual":[4000,1000,3000],"check":"control: refund limited by child cap","expected":[4000,1000,3000],"passed":true},{"actual":[4400,4400,0],"check":"control: fraction thereof","expected":[4400,4400,0],"passed":true},{"actual":[2000,0,1125],"check":"control: earned limit","expected":[2000,0,1125],"passed":true},{"actual":[6450,1000,5100],"check":"control: joint partial phase-out","expected":[6450,1000,5100],"passed":true}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"regression phaseout-floor 1\", \"actual\": [-500, -500, 0], \"expected\": [0, 0, 0], \"passed\": false}, {\"check\": \"regression phaseout-floor 2\", \"actual\": [-12500, -12500, 0], \"expected\": [0, 0, 0], \"passed\": false}, {\"check\": \"partial repair guard 1\", \"actual\": [500, 500, 0], \"expected\": [500, 500, 0], \"passed\": true}, {\"check\": \"partial repair guard 2\", \"actual\": [950, 500, 0], \"expected\": [950, 500, 0], \"passed\": true}, {\"check\": \"control: refund limited by child cap\", \"actual\": [4000, 1000, 3000], \"expected\": [4000, 1000, 3000], \"passed\": true}, {\"check\": \"control: fraction thereof\", \"actual\": [4400, 4400, 0], \"expected\": [4400, 4400, 0], \"passed\": true}, {\"check\": \"control: earned limit\", \"actual\": [2000, 0, 1125], \"expected\": [2000, 0, 1125], \"passed\": true}, {\"check\": \"control: joint partial phase-out\", \"actual\": [6450, 1000, 5100], \"expected\": [6450, 1000, 5100], \"passed\": true}], \"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."}}