{
 "jev": {
  "overall": {
   "n": 726,
   "accuracy": 0.854,
   "brier": 0.107
  },
  "by_role": {
   "code_reviewer": {
    "n": 180,
    "accuracy": 0.883,
    "brier": 0.099
   },
   "plan_reviewer": {
    "n": 60,
    "accuracy": 0.75,
    "brier": 0.18
   },
   "spec_reviewer": {
    "n": 178,
    "accuracy": 0.758,
    "brier": 0.158
   },
   "task_reviewer": {
    "n": 180,
    "accuracy": 0.922,
    "brier": 0.06
   },
   "test_reviewer": {
    "n": 128,
    "accuracy": 0.898,
    "brier": 0.078
   }
  },
  "by_kind": {
   "defect noul": {
    "n": 269,
    "accuracy": 0.833,
    "brier": 0.127
   },
   "per-test label": {
    "n": 128,
    "accuracy": 0.898,
    "brier": 0.078
   },
   "plan: any lens blocks": {
    "n": 60,
    "accuracy": 0.75,
    "brier": 0.18
   },
   "verdict": {
    "n": 269,
    "accuracy": 0.877,
    "brier": 0.084
   }
  },
  "errors": 0,
  "cost_usd": 0.0
 },
 "laya": {
  "overall": {
   "n": 726,
   "accuracy": 0.634,
   "brier": 0.249
  },
  "by_role": {
   "code_reviewer": {
    "n": 180,
    "accuracy": 0.678,
    "brier": 0.232
   },
   "plan_reviewer": {
    "n": 60,
    "accuracy": 0.333,
    "brier": 0.362
   },
   "spec_reviewer": {
    "n": 178,
    "accuracy": 0.674,
    "brier": 0.236
   },
   "task_reviewer": {
    "n": 180,
    "accuracy": 0.656,
    "brier": 0.235
   },
   "test_reviewer": {
    "n": 128,
    "accuracy": 0.625,
    "brier": 0.256
   }
  },
  "by_kind": {
   "defect noul": {
    "n": 269,
    "accuracy": 0.662,
    "brier": 0.237
   },
   "per-test label": {
    "n": 128,
    "accuracy": 0.625,
    "brier": 0.256
   },
   "plan: any lens blocks": {
    "n": 60,
    "accuracy": 0.333,
    "brier": 0.362
   },
   "verdict": {
    "n": 269,
    "accuracy": 0.677,
    "brier": 0.231
   }
  },
  "errors": 0,
  "cost_usd": 0.0
 },
 "claude": {
  "overall": {
   "n": 726,
   "accuracy": 0.915,
   "brier": 0.066
  },
  "by_role": {
   "code_reviewer": {
    "n": 180,
    "accuracy": 0.922,
    "brier": 0.067
   },
   "plan_reviewer": {
    "n": 60,
    "accuracy": 0.75,
    "brier": 0.18
   },
   "spec_reviewer": {
    "n": 178,
    "accuracy": 0.876,
    "brier": 0.076
   },
   "task_reviewer": {
    "n": 180,
    "accuracy": 0.972,
    "brier": 0.039
   },
   "test_reviewer": {
    "n": 128,
    "accuracy": 0.953,
    "brier": 0.037
   }
  },
  "by_kind": {
   "defect noul": {
    "n": 269,
    "accuracy": 0.944,
    "brier": 0.049
   },
   "per-test label": {
    "n": 128,
    "accuracy": 0.953,
    "brier": 0.037
   },
   "plan: any lens blocks": {
    "n": 60,
    "accuracy": 0.75,
    "brier": 0.18
   },
   "verdict": {
    "n": 269,
    "accuracy": 0.903,
    "brier": 0.071
   }
  },
  "errors": 0,
  "cost_usd": 194.26
 },
 "opus-5-5": {
  "overall": {
   "n": 726,
   "accuracy": 0.945,
   "brier": 0.051
  },
  "by_role": {
   "code_reviewer": {
    "n": 180,
    "accuracy": 0.933,
    "brier": 0.05
   },
   "plan_reviewer": {
    "n": 60,
    "accuracy": 0.733,
    "brier": 0.162
   },
   "spec_reviewer": {
    "n": 178,
    "accuracy": 0.972,
    "brier": 0.05
   },
   "task_reviewer": {
    "n": 180,
    "accuracy": 0.978,
    "brier": 0.029
   },
   "test_reviewer": {
    "n": 128,
    "accuracy": 0.977,
    "brier": 0.031
   }
  },
  "by_kind": {
   "defect noul": {
    "n": 269,
    "accuracy": 0.967,
    "brier": 0.039
   },
   "per-test label": {
    "n": 128,
    "accuracy": 0.977,
    "brier": 0.031
   },
   "plan: any lens blocks": {
    "n": 60,
    "accuracy": 0.733,
    "brier": 0.162
   },
   "verdict": {
    "n": 269,
    "accuracy": 0.955,
    "brier": 0.047
   }
  },
  "errors": 0,
  "cost_usd": 146.57
 },
 "sonnet-5-5": {
  "overall": {
   "n": 726,
   "accuracy": 0.858,
   "brier": 0.102
  },
  "by_role": {
   "code_reviewer": {
    "n": 180,
    "accuracy": 0.933,
    "brier": 0.046
   },
   "plan_reviewer": {
    "n": 60,
    "accuracy": 0.733,
    "brier": 0.144
   },
   "spec_reviewer": {
    "n": 178,
    "accuracy": 0.764,
    "brier": 0.177
   },
   "task_reviewer": {
    "n": 180,
    "accuracy": 0.939,
    "brier": 0.048
   },
   "test_reviewer": {
    "n": 128,
    "accuracy": 0.828,
    "brier": 0.133
   }
  },
  "by_kind": {
   "defect noul": {
    "n": 269,
    "accuracy": 0.9,
    "brier": 0.081
   },
   "per-test label": {
    "n": 128,
    "accuracy": 0.828,
    "brier": 0.133
   },
   "plan: any lens blocks": {
    "n": 60,
    "accuracy": 0.733,
    "brier": 0.144
   },
   "verdict": {
    "n": 269,
    "accuracy": 0.859,
    "brier": 0.099
   }
  },
  "errors": 0,
  "cost_usd": 73.52
 }
}