{
  "canary_drift": {
    "endpoint_max_range": {
      "deepseek-v4-flash-deepinfra-fp8": null,
      "minimax-m3-minimax-fp8": null,
      "moonshot-kimi-k27-inceptron-int4": null,
      "nvidia-nemotron35-coreweave-bf16": null,
      "qwen38-27b-chutes-fp8": null,
      "zai-glm53-zai-fp8": null
    },
    "epochs": {
      "end": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "failure_rate": null,
          "observed": 0
        },
        "minimax-m3-minimax-fp8": {
          "failure_rate": null,
          "observed": 0
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "failure_rate": null,
          "observed": 0
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "failure_rate": null,
          "observed": 0
        },
        "qwen38-27b-chutes-fp8": {
          "failure_rate": null,
          "observed": 0
        },
        "zai-glm53-zai-fp8": {
          "failure_rate": null,
          "observed": 0
        }
      },
      "midpoint": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "failure_rate": null,
          "observed": 0
        },
        "minimax-m3-minimax-fp8": {
          "failure_rate": null,
          "observed": 0
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "failure_rate": null,
          "observed": 0
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "failure_rate": null,
          "observed": 0
        },
        "qwen38-27b-chutes-fp8": {
          "failure_rate": null,
          "observed": 0
        },
        "zai-glm53-zai-fp8": {
          "failure_rate": null,
          "observed": 0
        }
      },
      "start": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "failure_rate": null,
          "observed": 0
        },
        "minimax-m3-minimax-fp8": {
          "failure_rate": null,
          "observed": 0
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "failure_rate": null,
          "observed": 0
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "failure_rate": null,
          "observed": 0
        },
        "qwen38-27b-chutes-fp8": {
          "failure_rate": null,
          "observed": 0
        },
        "zai-glm53-zai-fp8": {
          "failure_rate": null,
          "observed": 0
        }
      }
    }
  },
  "cell_states": {
    "accepted": 3585,
    "ambiguous": 3,
    "infrastructure_error": 12
  },
  "conditions": {
    "A_independent": {
      "complete_fraction": 0.96,
      "complete_items": 96,
      "domain": {
        "arithmetic": {
          "failure_rate": 0.36752136752136755,
          "observed": 117
        },
        "code": {
          "failure_rate": 0.6583333333333333,
          "observed": 120
        },
        "constraint": {
          "failure_rate": 0.43333333333333335,
          "observed": 120
        },
        "extraction": {
          "failure_rate": 0.5126050420168067,
          "observed": 119
        },
        "tool_call": {
          "failure_rate": 0.9833333333333333,
          "observed": 120
        }
      },
      "endpoint": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.54,
          "failures": 54,
          "observed": 100,
          "passes": 46,
          "planned": 100
        },
        "minimax-m3-minimax-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.57,
          "failures": 57,
          "observed": 100,
          "passes": 43,
          "planned": 100
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "complete_fraction": 0.99,
          "failure_rate": 0.35353535353535354,
          "failures": 35,
          "observed": 99,
          "passes": 64,
          "planned": 100
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "complete_fraction": 0.98,
          "failure_rate": 0.9897959183673469,
          "failures": 97,
          "observed": 98,
          "passes": 1,
          "planned": 100
        },
        "qwen38-27b-chutes-fp8": {
          "complete_fraction": 0.99,
          "failure_rate": 0.7676767676767676,
          "failures": 76,
          "observed": 99,
          "passes": 23,
          "planned": 100
        },
        "zai-glm53-zai-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.34,
          "failures": 34,
          "observed": 100,
          "passes": 66,
          "planned": 100
        }
      },
      "metrics": {
        "all_failure_rate": 0.1875,
        "covariance_diversification_index": 2.345839977670474,
        "dependence_inflation": 2.2881421027364377,
        "diagonal_variance": 0.03172210038986355,
        "endpoints": 6,
        "fleet_variance": 0.07258467348927874,
        "items": 96,
        "majority_failure_rate": 0.75,
        "mean_loss": 0.6059027777777778,
        "off_diagonal_covariance": 0.0408625730994152
      },
      "missingness_bounds": {
        "all_failure_high": 0.18,
        "all_failure_low": 0.18,
        "items": 100,
        "majority_failure_high": 0.73,
        "majority_failure_low": 0.72,
        "mean_loss_high": 0.595,
        "mean_loss_low": 0.5883333333333334,
        "missing_cells": 4
      },
      "nondegenerate_endpoint_pairs": 15,
      "nondegenerate_endpoints": 6,
      "universal_pass_rate": 0.010416666666666666
    },
    "A_shared": {
      "complete_fraction": 0.98,
      "complete_items": 98,
      "domain": {
        "arithmetic": {
          "failure_rate": 0.3416666666666667,
          "observed": 120
        },
        "code": {
          "failure_rate": 0.6218487394957983,
          "observed": 119
        },
        "constraint": {
          "failure_rate": 0.44166666666666665,
          "observed": 120
        },
        "extraction": {
          "failure_rate": 0.49166666666666664,
          "observed": 120
        },
        "tool_call": {
          "failure_rate": 0.957983193277311,
          "observed": 119
        }
      },
      "endpoint": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.47,
          "failures": 47,
          "observed": 100,
          "passes": 53,
          "planned": 100
        },
        "minimax-m3-minimax-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.57,
          "failures": 57,
          "observed": 100,
          "passes": 43,
          "planned": 100
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "complete_fraction": 0.99,
          "failure_rate": 0.3333333333333333,
          "failures": 33,
          "observed": 99,
          "passes": 66,
          "planned": 100
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "complete_fraction": 0.99,
          "failure_rate": 0.9797979797979798,
          "failures": 97,
          "observed": 99,
          "passes": 2,
          "planned": 100
        },
        "qwen38-27b-chutes-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.72,
          "failures": 72,
          "observed": 100,
          "passes": 28,
          "planned": 100
        },
        "zai-glm53-zai-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.35,
          "failures": 35,
          "observed": 100,
          "passes": 65,
          "planned": 100
        }
      },
      "metrics": {
        "all_failure_rate": 0.16326530612244897,
        "covariance_diversification_index": 2.271666055125524,
        "dependence_inflation": 2.423152270703473,
        "diagonal_variance": 0.03281553171096617,
        "endpoints": 6,
        "fleet_variance": 0.0795170301797695,
        "items": 98,
        "majority_failure_rate": 0.673469387755102,
        "mean_loss": 0.5680272108843537,
        "off_diagonal_covariance": 0.046701498468803324
      },
      "missingness_bounds": {
        "all_failure_high": 0.16,
        "all_failure_low": 0.16,
        "items": 100,
        "majority_failure_high": 0.68,
        "majority_failure_low": 0.68,
        "mean_loss_high": 0.5716666666666667,
        "mean_loss_low": 0.5683333333333334,
        "missing_cells": 2
      },
      "nondegenerate_endpoint_pairs": 15,
      "nondegenerate_endpoints": 6,
      "universal_pass_rate": 0.02040816326530612
    },
    "E_independent": {
      "complete_fraction": 0.99,
      "complete_items": 99,
      "domain": {
        "arithmetic": {
          "failure_rate": 0.31666666666666665,
          "observed": 120
        },
        "code": {
          "failure_rate": 0.6,
          "observed": 120
        },
        "constraint": {
          "failure_rate": 0.425,
          "observed": 120
        },
        "extraction": {
          "failure_rate": 0.5462184873949579,
          "observed": 119
        },
        "tool_call": {
          "failure_rate": 0.975,
          "observed": 120
        }
      },
      "endpoint": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.38,
          "failures": 38,
          "observed": 100,
          "passes": 62,
          "planned": 100
        },
        "minimax-m3-minimax-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.54,
          "failures": 54,
          "observed": 100,
          "passes": 46,
          "planned": 100
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "complete_fraction": 0.99,
          "failure_rate": 0.3838383838383838,
          "failures": 38,
          "observed": 99,
          "passes": 61,
          "planned": 100
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "complete_fraction": 1.0,
          "failure_rate": 1.0,
          "failures": 100,
          "observed": 100,
          "passes": 0,
          "planned": 100
        },
        "qwen38-27b-chutes-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.81,
          "failures": 81,
          "observed": 100,
          "passes": 19,
          "planned": 100
        },
        "zai-glm53-zai-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.32,
          "failures": 32,
          "observed": 100,
          "passes": 68,
          "planned": 100
        }
      },
      "metrics": {
        "all_failure_rate": 0.1717171717171717,
        "covariance_diversification_index": 1.9931818229536298,
        "dependence_inflation": 2.492270441422983,
        "diagonal_variance": 0.030743947410614078,
        "endpoints": 6,
        "fleet_variance": 0.07662223138413612,
        "items": 99,
        "majority_failure_rate": 0.6464646464646465,
        "mean_loss": 0.5723905723905723,
        "off_diagonal_covariance": 0.04587828397352205
      },
      "missingness_bounds": {
        "all_failure_high": 0.17,
        "all_failure_low": 0.17,
        "items": 100,
        "majority_failure_high": 0.65,
        "majority_failure_low": 0.65,
        "mean_loss_high": 0.5733333333333334,
        "mean_loss_low": 0.5716666666666667,
        "missing_cells": 1
      },
      "nondegenerate_endpoint_pairs": 10,
      "nondegenerate_endpoints": 5,
      "universal_pass_rate": 0.0
    },
    "E_shared": {
      "complete_fraction": 0.98,
      "complete_items": 98,
      "domain": {
        "arithmetic": {
          "failure_rate": 0.2916666666666667,
          "observed": 120
        },
        "code": {
          "failure_rate": 0.6833333333333333,
          "observed": 120
        },
        "constraint": {
          "failure_rate": 0.42016806722689076,
          "observed": 119
        },
        "extraction": {
          "failure_rate": 0.5042016806722689,
          "observed": 119
        },
        "tool_call": {
          "failure_rate": 0.9416666666666667,
          "observed": 120
        }
      },
      "endpoint": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.39,
          "failures": 39,
          "observed": 100,
          "passes": 61,
          "planned": 100
        },
        "minimax-m3-minimax-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.54,
          "failures": 54,
          "observed": 100,
          "passes": 46,
          "planned": 100
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "complete_fraction": 0.99,
          "failure_rate": 0.36363636363636365,
          "failures": 36,
          "observed": 99,
          "passes": 63,
          "planned": 100
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "complete_fraction": 0.99,
          "failure_rate": 1.0,
          "failures": 99,
          "observed": 99,
          "passes": 0,
          "planned": 100
        },
        "qwen38-27b-chutes-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.8,
          "failures": 80,
          "observed": 100,
          "passes": 20,
          "planned": 100
        },
        "zai-glm53-zai-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.32,
          "failures": 32,
          "observed": 100,
          "passes": 68,
          "planned": 100
        }
      },
      "metrics": {
        "all_failure_rate": 0.16326530612244897,
        "covariance_diversification_index": 1.9722696528465928,
        "dependence_inflation": 2.522040003791828,
        "diagonal_variance": 0.030825560464735724,
        "endpoints": 6,
        "fleet_variance": 0.07774329663136731,
        "items": 98,
        "majority_failure_rate": 0.673469387755102,
        "mean_loss": 0.5663265306122449,
        "off_diagonal_covariance": 0.046917736166631585
      },
      "missingness_bounds": {
        "all_failure_high": 0.16,
        "all_failure_low": 0.16,
        "items": 100,
        "majority_failure_high": 0.68,
        "majority_failure_low": 0.68,
        "mean_loss_high": 0.57,
        "mean_loss_low": 0.5666666666666667,
        "missing_cells": 2
      },
      "nondegenerate_endpoint_pairs": 10,
      "nondegenerate_endpoints": 5,
      "universal_pass_rate": 0.0
    },
    "N": {
      "complete_fraction": 0.97,
      "complete_items": 97,
      "domain": {
        "arithmetic": {
          "failure_rate": 0.211864406779661,
          "observed": 118
        },
        "code": {
          "failure_rate": 0.5166666666666667,
          "observed": 120
        },
        "constraint": {
          "failure_rate": 0.425,
          "observed": 120
        },
        "extraction": {
          "failure_rate": 0.6416666666666667,
          "observed": 120
        },
        "tool_call": {
          "failure_rate": 0.957983193277311,
          "observed": 119
        }
      },
      "endpoint": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.4,
          "failures": 40,
          "observed": 100,
          "passes": 60,
          "planned": 100
        },
        "minimax-m3-minimax-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.49,
          "failures": 49,
          "observed": 100,
          "passes": 51,
          "planned": 100
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "complete_fraction": 1.0,
          "failure_rate": 0.4,
          "failures": 40,
          "observed": 100,
          "passes": 60,
          "planned": 100
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "complete_fraction": 0.98,
          "failure_rate": 0.9693877551020408,
          "failures": 95,
          "observed": 98,
          "passes": 3,
          "planned": 100
        },
        "qwen38-27b-chutes-fp8": {
          "complete_fraction": 0.99,
          "failure_rate": 0.6262626262626263,
          "failures": 62,
          "observed": 99,
          "passes": 37,
          "planned": 100
        },
        "zai-glm53-zai-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.43,
          "failures": 43,
          "observed": 100,
          "passes": 57,
          "planned": 100
        }
      },
      "metrics": {
        "all_failure_rate": 0.18556701030927836,
        "covariance_diversification_index": 1.991473280684366,
        "dependence_inflation": 2.8067226890756305,
        "diagonal_variance": 0.03478784841542573,
        "endpoints": 6,
        "fleet_variance": 0.09763984345169913,
        "items": 97,
        "majority_failure_rate": 0.5979381443298969,
        "mean_loss": 0.5549828178694158,
        "off_diagonal_covariance": 0.0628519950362734
      },
      "missingness_bounds": {
        "all_failure_high": 0.18,
        "all_failure_low": 0.18,
        "items": 100,
        "majority_failure_high": 0.6,
        "majority_failure_low": 0.59,
        "mean_loss_high": 0.5533333333333333,
        "mean_loss_low": 0.5483333333333333,
        "missing_cells": 3
      },
      "nondegenerate_endpoint_pairs": 15,
      "nondegenerate_endpoints": 6,
      "universal_pass_rate": 0.030927835051546393
    },
    "O": {
      "complete_fraction": 0.97,
      "complete_items": 97,
      "domain": {
        "arithmetic": {
          "failure_rate": 0.22033898305084745,
          "observed": 118
        },
        "code": {
          "failure_rate": 0.4583333333333333,
          "observed": 120
        },
        "constraint": {
          "failure_rate": 0.4,
          "observed": 120
        },
        "extraction": {
          "failure_rate": 0.5546218487394958,
          "observed": 119
        },
        "tool_call": {
          "failure_rate": 0.9416666666666667,
          "observed": 120
        }
      },
      "endpoint": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.39,
          "failures": 39,
          "observed": 100,
          "passes": 61,
          "planned": 100
        },
        "minimax-m3-minimax-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.42,
          "failures": 42,
          "observed": 100,
          "passes": 58,
          "planned": 100
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "complete_fraction": 0.99,
          "failure_rate": 0.37373737373737376,
          "failures": 37,
          "observed": 99,
          "passes": 62,
          "planned": 100
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "complete_fraction": 0.99,
          "failure_rate": 0.9797979797979798,
          "failures": 97,
          "observed": 99,
          "passes": 2,
          "planned": 100
        },
        "qwen38-27b-chutes-fp8": {
          "complete_fraction": 0.99,
          "failure_rate": 0.5555555555555556,
          "failures": 55,
          "observed": 99,
          "passes": 44,
          "planned": 100
        },
        "zai-glm53-zai-fp8": {
          "complete_fraction": 1.0,
          "failure_rate": 0.38,
          "failures": 38,
          "observed": 100,
          "passes": 62,
          "planned": 100
        }
      },
      "metrics": {
        "all_failure_rate": 0.16494845360824742,
        "covariance_diversification_index": 1.9730026556777485,
        "dependence_inflation": 2.7890178416767717,
        "diagonal_variance": 0.0344418193967163,
        "endpoints": 6,
        "fleet_variance": 0.09605884879725085,
        "items": 97,
        "majority_failure_rate": 0.5154639175257731,
        "mean_loss": 0.5240549828178693,
        "off_diagonal_covariance": 0.061617029400534554
      },
      "missingness_bounds": {
        "all_failure_high": 0.16,
        "all_failure_low": 0.16,
        "items": 100,
        "majority_failure_high": 0.51,
        "majority_failure_low": 0.5,
        "mean_loss_high": 0.5183333333333333,
        "mean_loss_low": 0.5133333333333333,
        "missing_cells": 3
      },
      "nondegenerate_endpoint_pairs": 15,
      "nondegenerate_endpoints": 6,
      "universal_pass_rate": 0.020618556701030927
    }
  },
  "difficulty_residual_sensitivity": {
    "contrasts": {
      "evidence_corruption": {
        "independent": {
          "dependence_inflation": 0.8436064104106854,
          "diagonal_variance": 0.022288135768146866,
          "fleet_variance": 0.018802414210112382
        },
        "log_dependence_inflation_ratio": 0.5053443449092131,
        "matched_items": 94,
        "shared": {
          "dependence_inflation": 1.3983250302314718,
          "diagonal_variance": 0.027827113251900282,
          "fleet_variance": 0.03891134897921805
        }
      },
      "instruction_anchor": {
        "independent": {
          "dependence_inflation": 1.1404935367418434,
          "diagonal_variance": 0.025850682300176557,
          "fleet_variance": 0.029482536083718133
        },
        "log_dependence_inflation_ratio": 0.050078152349232456,
        "matched_items": 93,
        "shared": {
          "dependence_inflation": 1.199061596599303,
          "diagonal_variance": 0.02747003342728644,
          "fleet_variance": 0.032938262139958305
        }
      }
    },
    "method": "five-fold item-family cross-fit; leave-endpoint-out O difficulty plus frozen domain and predicted-band features; ridge logistic endpoint calibration",
    "prediction_support_fraction": 0.995,
    "schema_version": "cmfc.common-shock-residual-sensitivity.v1"
  },
  "endpoints": [
    "deepseek-v4-flash-deepinfra-fp8",
    "qwen38-27b-chutes-fp8",
    "zai-glm53-zai-fp8",
    "minimax-m3-minimax-fp8",
    "moonshot-kimi-k27-inceptron-int4",
    "nvidia-nemotron35-coreweave-bf16"
  ],
  "execution_id": "common-shock-contemporary-v8-20260826",
  "metric_boundaries": {
    "absolute_volatility_ratio": "shared fleet SD divided by independent fleet SD",
    "capital_claim": "none",
    "dependence_inflation": "fleet variance divided by condition-specific diagonal variance",
    "n_eff": "secondary benchmark-relative covariance-diversification index"
  },
  "neutral_equivalence": {
    "dependence_inflation_ratio_N_over_O": 1.0121342295121691,
    "dependence_ratio_margin": [
      0.9,
      1.1
    ],
    "dependence_ratio_within_margin": true,
    "matched_items": 95,
    "mean_difference_O_minus_N": -0.033333333333333215,
    "mean_margin": [
      -0.03,
      0.03
    ],
    "mean_within_margin": false
  },
  "planned_base_items": 100,
  "primary_contrasts": {
    "evidence_corruption": {
      "absolute_volatility_ratio": 1.0062903351186923,
      "all_failure_difference": -0.010309278350515455,
      "dependence_inflation_ratio": 1.0071561093030503,
      "independent_condition": "E_independent",
      "log_dependence_inflation_ratio": 0.0071306258555374384,
      "majority_failure_difference": 0.020618556701030855,
      "matched_fraction": 0.97,
      "matched_items": 97,
      "mean_loss_difference": -0.0068728522336770626,
      "mean_loss_difference_bounds": {
        "high": -0.0016666666666667052,
        "low": -0.00666666666666671
      },
      "per_endpoint_induced_failure": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "independent_failure_given_shared_pass": 0.22033898305084745,
          "shared_failure_given_independent_pass": 0.22033898305084745
        },
        "minimax-m3-minimax-fp8": {
          "independent_failure_given_shared_pass": 0.3695652173913043,
          "shared_failure_given_independent_pass": 0.3409090909090909
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "independent_failure_given_shared_pass": 0.20967741935483872,
          "shared_failure_given_independent_pass": 0.18333333333333332
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "independent_failure_given_shared_pass": null,
          "shared_failure_given_independent_pass": null
        },
        "qwen38-27b-chutes-fp8": {
          "independent_failure_given_shared_pass": 0.35,
          "shared_failure_given_independent_pass": 0.3157894736842105
        },
        "zai-glm53-zai-fp8": {
          "independent_failure_given_shared_pass": 0.06153846153846154,
          "shared_failure_given_independent_pass": 0.07575757575757576
        }
      },
      "planned_items": 100,
      "shared_condition": "E_shared",
      "support_by_domain": {
        "arithmetic": {
          "matched_items": 20,
          "planned_items": 20
        },
        "code": {
          "matched_items": 20,
          "planned_items": 20
        },
        "constraint": {
          "matched_items": 19,
          "planned_items": 20
        },
        "extraction": {
          "matched_items": 18,
          "planned_items": 20
        },
        "tool_call": {
          "matched_items": 20,
          "planned_items": 20
        }
      },
      "support_classification": "confirmatory_quality"
    },
    "instruction_anchor": {
      "absolute_volatility_ratio": 1.0304171508971272,
      "all_failure_difference": -0.010638297872340441,
      "dependence_inflation_ratio": 1.0329067823678943,
      "independent_condition": "A_independent",
      "log_dependence_inflation_ratio": 0.03237694634435022,
      "majority_failure_difference": -0.04255319148936165,
      "matched_fraction": 0.94,
      "matched_items": 94,
      "mean_loss_difference": -0.015957446808510634,
      "mean_loss_difference_bounds": {
        "high": -0.01666666666666672,
        "low": -0.026666666666666616
      },
      "per_endpoint_induced_failure": {
        "deepseek-v4-flash-deepinfra-fp8": {
          "independent_failure_given_shared_pass": 0.22916666666666666,
          "shared_failure_given_independent_pass": 0.13953488372093023
        },
        "minimax-m3-minimax-fp8": {
          "independent_failure_given_shared_pass": 0.358974358974359,
          "shared_failure_given_independent_pass": 0.358974358974359
        },
        "moonshot-kimi-k27-inceptron-int4": {
          "independent_failure_given_shared_pass": 0.13114754098360656,
          "shared_failure_given_independent_pass": 0.11666666666666667
        },
        "nvidia-nemotron35-coreweave-bf16": {
          "independent_failure_given_shared_pass": 0.0,
          "shared_failure_given_independent_pass": 0.0
        },
        "qwen38-27b-chutes-fp8": {
          "independent_failure_given_shared_pass": 0.34615384615384615,
          "shared_failure_given_independent_pass": 0.19047619047619047
        },
        "zai-glm53-zai-fp8": {
          "independent_failure_given_shared_pass": 0.15,
          "shared_failure_given_independent_pass": 0.1774193548387097
        }
      },
      "planned_items": 100,
      "shared_condition": "A_shared",
      "support_by_domain": {
        "arithmetic": {
          "matched_items": 17,
          "planned_items": 20
        },
        "code": {
          "matched_items": 19,
          "planned_items": 20
        },
        "constraint": {
          "matched_items": 20,
          "planned_items": 20
        },
        "extraction": {
          "matched_items": 19,
          "planned_items": 20
        },
        "tool_call": {
          "matched_items": 19,
          "planned_items": 20
        }
      },
      "support_classification": "confirmatory_quality"
    }
  },
  "provider_availability": {
    "ambiguous_events": 3,
    "boundary": "retry/missing delivery diagnostics are not recoded as model failures or pooled into failure dependence",
    "endpoint": {
      "deepseek-v4-flash-deepinfra-fp8": {
        "accepted": 600,
        "error_events": 2,
        "missing_cells": 0,
        "planned": 600,
        "retry_cells": 2,
        "retry_failure_phi_on_accepted": -0.05005920030528664
      },
      "minimax-m3-minimax-fp8": {
        "accepted": 600,
        "error_events": 0,
        "missing_cells": 0,
        "planned": 600,
        "retry_cells": 0,
        "retry_failure_phi_on_accepted": null
      },
      "moonshot-kimi-k27-inceptron-int4": {
        "accepted": 595,
        "error_events": 79,
        "missing_cells": 5,
        "planned": 600,
        "retry_cells": 74,
        "retry_failure_phi_on_accepted": -0.015200841354147617
      },
      "nvidia-nemotron35-coreweave-bf16": {
        "accepted": 593,
        "error_events": 33,
        "missing_cells": 7,
        "planned": 600,
        "retry_cells": 26,
        "retry_failure_phi_on_accepted": -0.06172810095664727
      },
      "qwen38-27b-chutes-fp8": {
        "accepted": 597,
        "error_events": 0,
        "missing_cells": 3,
        "planned": 600,
        "retry_cells": 0,
        "retry_failure_phi_on_accepted": null
      },
      "zai-glm53-zai-fp8": {
        "accepted": 600,
        "error_events": 0,
        "missing_cells": 0,
        "planned": 600,
        "retry_cells": 0,
        "retry_failure_phi_on_accepted": null
      }
    },
    "quarantine_events": 0,
    "schema_version": "cmfc.common-shock-availability-diagnostic.v1",
    "terminal_error_types": {
      "HTTP_429": 114,
      "HTTP_504": 3
    }
  },
  "schema_version": "cmfc.common-shock-analysis.v2",
  "simultaneous_inference": {
    "confidence": 0.95,
    "critical_value": 2.2325827651498056,
    "finite_joint_draws": 9999,
    "intervals": {
      "evidence_corruption": {
        "estimate": 0.0071306258555374384,
        "finite_draws": 9999,
        "high": 0.13267361537953723,
        "low": -0.11841236366846235,
        "standard_error": 0.056232177137484934
      },
      "instruction_anchor": {
        "estimate": 0.03237694634435022,
        "finite_draws": 9999,
        "high": 0.14265590403065304,
        "low": -0.0779020113419526,
        "standard_error": 0.049395238289812356
      }
    },
    "method": "studentized max-absolute-statistic item-family-by-stratum intersection cluster bootstrap; singleton-cluster strata are self-representing certainty strata",
    "resamples": 9999,
    "schema_version": "cmfc.common-shock-simultaneous-inference.v1",
    "seed": 20260823,
    "tail_probability_monte_carlo_se": 0.002179558452417543
  },
  "subgroup_sensitivity": {
    "heterogeneity": {
      "A_shared_variant": {
        "a_shared-00": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1205720941473682,
            "all_failure_difference": -0.1,
            "dependence_inflation_ratio": 1.3178442844284433,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.27599728375040905,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.01666666666666672,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9507082380237135,
            "all_failure_difference": -0.1,
            "dependence_inflation_ratio": 0.9323886639676112,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.07000552980133781,
            "majority_failure_difference": -0.09999999999999998,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.01666666666666672,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "a_shared-01": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.7169509687480624,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.6043956043956046,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.5035263212843788,
            "majority_failure_difference": 0.4,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.033333333333333215,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.8933642943784517,
            "all_failure_difference": -0.2,
            "dependence_inflation_ratio": 0.9712723524402816,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.029148363469572006,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.08333333333333337,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "a_shared-02": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.044465935734187,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.0047846889952154,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.004773278752657812,
            "majority_failure_difference": -0.2222222222222222,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.11111111111111105,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.0359395405656244,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.2382739212007503,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.21371841085462656,
            "majority_failure_difference": -0.11111111111111116,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.03703703703703698,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "a_shared-03": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1095100982991764,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.1067820229938448,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.10145672647296211,
            "majority_failure_difference": 0.2,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.016666666666666663,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9483517644747597,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.89937106918239,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.10605957196032421,
            "majority_failure_difference": 0.125,
            "matched_fraction": 0.8,
            "matched_items": 8,
            "mean_loss_difference": 0.04166666666666663,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "a_shared-04": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1125816095180783,
            "all_failure_difference": 0.1,
            "dependence_inflation_ratio": 1.2878514878514877,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.25297531657096056,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.06666666666666665,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9999999999999999,
            "all_failure_difference": 0.1111111111111111,
            "dependence_inflation_ratio": 0.9130434782608696,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.09097177820572663,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.01851851851851849,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "a_shared-05": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1451966686277366,
            "all_failure_difference": 0.09999999999999998,
            "dependence_inflation_ratio": 1.241217798594848,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.21609299331754317,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.06666666666666676,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9621073524408894,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8765009704905088,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.13181746738756606,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.0,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "a_shared-06": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.941340000446853,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8222564201211888,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.1957029859517365,
            "majority_failure_difference": 0.10000000000000009,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.033333333333333326,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.1435437497937313,
            "all_failure_difference": 0.10000000000000003,
            "dependence_inflation_ratio": 1.148494983277592,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.13845237507268868,
            "majority_failure_difference": -0.10000000000000009,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.033333333333333326,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "a_shared-07": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.806946584785929,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8879492600422834,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.11884067721451891,
            "majority_failure_difference": -0.11111111111111116,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.055555555555555525,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.1055415967851332,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.94017094017094,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.06169356900534002,
            "majority_failure_difference": -0.11111111111111116,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.018518518518518545,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "a_shared-08": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.7578418860978352,
            "all_failure_difference": -0.2222222222222222,
            "dependence_inflation_ratio": 0.6317567567567566,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.45925083746947404,
            "majority_failure_difference": 0.2222222222222222,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": 0.037037037037037035,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.160181348430354,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.3226117045283585,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.2796083458551045,
            "majority_failure_difference": -0.09999999999999998,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.033333333333333326,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "a_shared-09": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1578536176909409,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.3285472972972974,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.2840860876143269,
            "majority_failure_difference": -0.4,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.1499999999999999,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.985329278164293,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8737864077669903,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.13491931789937064,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": 0.05555555555555558,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        }
      },
      "E_shared_variant": {
        "e_shared-00": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1451966686277366,
            "all_failure_difference": 0.09999999999999998,
            "dependence_inflation_ratio": 1.241217798594848,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.21609299331754317,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.06666666666666676,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9621073524408894,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8765009704905088,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.13181746738756606,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.0,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "e_shared-01": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.941340000446853,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8222564201211888,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.1957029859517365,
            "majority_failure_difference": 0.10000000000000009,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.033333333333333326,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.1435437497937313,
            "all_failure_difference": 0.10000000000000003,
            "dependence_inflation_ratio": 1.148494983277592,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.13845237507268868,
            "majority_failure_difference": -0.10000000000000009,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.033333333333333326,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "e_shared-02": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.806946584785929,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8879492600422834,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.11884067721451891,
            "majority_failure_difference": -0.11111111111111116,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.055555555555555525,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.1055415967851332,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.94017094017094,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.06169356900534002,
            "majority_failure_difference": -0.11111111111111116,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.018518518518518545,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "e_shared-03": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.7578418860978352,
            "all_failure_difference": -0.2222222222222222,
            "dependence_inflation_ratio": 0.6317567567567566,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.45925083746947404,
            "majority_failure_difference": 0.2222222222222222,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": 0.037037037037037035,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.160181348430354,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.3226117045283585,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.2796083458551045,
            "majority_failure_difference": -0.09999999999999998,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.033333333333333326,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "e_shared-04": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1578536176909409,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.3285472972972974,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.2840860876143269,
            "majority_failure_difference": -0.4,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.1499999999999999,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.985329278164293,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8737864077669903,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.13491931789937064,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": 0.05555555555555558,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "e_shared-05": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1205720941473682,
            "all_failure_difference": -0.1,
            "dependence_inflation_ratio": 1.3178442844284433,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.27599728375040905,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.01666666666666672,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9507082380237135,
            "all_failure_difference": -0.1,
            "dependence_inflation_ratio": 0.9323886639676112,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.07000552980133781,
            "majority_failure_difference": -0.09999999999999998,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.01666666666666672,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "e_shared-06": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.7169509687480624,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.6043956043956046,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.5035263212843788,
            "majority_failure_difference": 0.4,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.033333333333333215,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.8933642943784517,
            "all_failure_difference": -0.2,
            "dependence_inflation_ratio": 0.9712723524402816,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.029148363469572006,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.08333333333333337,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "e_shared-07": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.044465935734187,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.0047846889952154,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.004773278752657812,
            "majority_failure_difference": -0.2222222222222222,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.11111111111111105,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.0359395405656244,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.2382739212007503,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.21371841085462656,
            "majority_failure_difference": -0.11111111111111116,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.03703703703703698,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "e_shared-08": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1095100982991764,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.1067820229938448,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.10145672647296211,
            "majority_failure_difference": 0.2,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.016666666666666663,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9483517644747597,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.89937106918239,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.10605957196032421,
            "majority_failure_difference": 0.125,
            "matched_fraction": 0.8,
            "matched_items": 8,
            "mean_loss_difference": 0.04166666666666663,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "e_shared-09": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1125816095180783,
            "all_failure_difference": 0.1,
            "dependence_inflation_ratio": 1.2878514878514877,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.25297531657096056,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.06666666666666665,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9999999999999999,
            "all_failure_difference": 0.1111111111111111,
            "dependence_inflation_ratio": 0.9130434782608696,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.09097177820572663,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": -0.01851851851851849,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        }
      },
      "domain": {
        "arithmetic": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.8642079297900186,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8312281255756427,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.18485100245529254,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 20,
            "mean_loss_difference": -0.025000000000000022,
            "planned_items": 20,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9567180599602327,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8879867762166368,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.11879842774878022,
            "majority_failure_difference": -0.05882352941176472,
            "matched_fraction": 0.85,
            "matched_items": 17,
            "mean_loss_difference": -0.029411764705882304,
            "planned_items": 20,
            "shared_condition": "A_shared"
          }
        },
        "code": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.9730408576407765,
            "all_failure_difference": 0.05,
            "dependence_inflation_ratio": 1.191561731075409,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.17526482576172453,
            "majority_failure_difference": 0.15000000000000002,
            "matched_fraction": 1.0,
            "matched_items": 20,
            "mean_loss_difference": 0.08333333333333337,
            "planned_items": 20,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.6684674955730434,
            "all_failure_difference": 0.05263157894736842,
            "dependence_inflation_ratio": 2.027495703796281,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.7068013880391099,
            "majority_failure_difference": -0.1578947368421053,
            "matched_fraction": 0.95,
            "matched_items": 19,
            "mean_loss_difference": -0.02631578947368418,
            "planned_items": 20,
            "shared_condition": "A_shared"
          }
        },
        "constraint": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.0,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.1818181818181817,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.16705408466316607,
            "majority_failure_difference": -0.05263157894736842,
            "matched_fraction": 0.95,
            "matched_items": 19,
            "mean_loss_difference": -0.02631578947368418,
            "planned_items": 20,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.1057581785477724,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.2749533352981628,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.24290957817445485,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 20,
            "mean_loss_difference": 0.008333333333333304,
            "planned_items": 20,
            "shared_condition": "A_shared"
          }
        },
        "extraction": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.9486832980505139,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.9356435643564357,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.06652068234156232,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.9,
            "matched_items": 18,
            "mean_loss_difference": -0.03703703703703698,
            "planned_items": 20,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9039935293326326,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8834641092705612,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.12390461139721266,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.95,
            "matched_items": 19,
            "mean_loss_difference": -0.01754385964912275,
            "planned_items": 20,
            "shared_condition": "A_shared"
          }
        },
        "tool_call": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.8311038136792213,
            "all_failure_difference": -0.09999999999999998,
            "dependence_inflation_ratio": 1.5240641711229947,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.421380563414064,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 20,
            "mean_loss_difference": -0.033333333333333326,
            "planned_items": 20,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.3284223283101433,
            "all_failure_difference": -0.10526315789473684,
            "dependence_inflation_ratio": 0.9999999999999999,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -1.1102230246251565e-16,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.95,
            "matched_items": 19,
            "mean_loss_difference": -0.01754385964912275,
            "planned_items": 20,
            "shared_condition": "A_shared"
          }
        }
      },
      "predicted_difficulty_band": {
        "tier-1": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.9168747682531899,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.9161031276415893,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.08762633588363504,
            "majority_failure_difference": 0.23529411764705876,
            "matched_fraction": 0.9444444444444444,
            "matched_items": 17,
            "mean_loss_difference": 0.08823529411764708,
            "planned_items": 18,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.0031555774234635,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.9759288641512316,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.024365580316243116,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.8888888888888888,
            "matched_items": 16,
            "mean_loss_difference": 0.03125,
            "planned_items": 18,
            "shared_condition": "A_shared"
          }
        },
        "tier-2": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.976436481196358,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.9717047807595817,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.0287032441533939,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 18,
            "mean_loss_difference": -0.018518518518518434,
            "planned_items": 18,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.0131488480155786,
            "all_failure_difference": 0.058823529411764705,
            "dependence_inflation_ratio": 0.8524925224327018,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.15959084120287287,
            "majority_failure_difference": 0.05882352941176472,
            "matched_fraction": 0.9444444444444444,
            "matched_items": 17,
            "mean_loss_difference": 0.039215686274509776,
            "planned_items": 18,
            "shared_condition": "A_shared"
          }
        },
        "tier-3": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1383209775310843,
            "all_failure_difference": 0.047619047619047616,
            "dependence_inflation_ratio": 1.1686191917862312,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.15582287389075877,
            "majority_failure_difference": -0.04761904761904756,
            "matched_fraction": 1.0,
            "matched_items": 21,
            "mean_loss_difference": 0.023809523809523836,
            "planned_items": 21,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.030427876755769,
            "all_failure_difference": -0.05,
            "dependence_inflation_ratio": 1.1347284373080633,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.12639336002193466,
            "majority_failure_difference": -0.20000000000000007,
            "matched_fraction": 0.9523809523809523,
            "matched_items": 20,
            "mean_loss_difference": -0.09166666666666656,
            "planned_items": 21,
            "shared_condition": "A_shared"
          }
        },
        "tier-4": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.084442471451489,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.111009593421654,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.10526914556319489,
            "majority_failure_difference": 0.03846153846153855,
            "matched_fraction": 0.9285714285714286,
            "matched_items": 26,
            "mean_loss_difference": 0.0,
            "planned_items": 28,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.0967264876732232,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.2043490770979297,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.18593923600917703,
            "majority_failure_difference": -0.038461538461538436,
            "matched_fraction": 0.9285714285714286,
            "matched_items": 26,
            "mean_loss_difference": -0.019230769230769162,
            "planned_items": 28,
            "shared_condition": "A_shared"
          }
        },
        "tier-5": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.1249142628509214,
            "all_failure_difference": -0.13333333333333336,
            "dependence_inflation_ratio": 1.1863425925925928,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.1708751227684531,
            "majority_failure_difference": -0.1333333333333333,
            "matched_fraction": 1.0,
            "matched_items": 15,
            "mean_loss_difference": -0.15555555555555545,
            "planned_items": 15,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9530986027504633,
            "all_failure_difference": -0.06666666666666665,
            "dependence_inflation_ratio": 0.9000630296239231,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.10529048519454627,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 15,
            "mean_loss_difference": -0.022222222222222254,
            "planned_items": 15,
            "shared_condition": "A_shared"
          }
        }
      },
      "template_family_id": {
        "cs-v7-arithmetic-lower-tier-batch-reconciliation": {
          "evidence_corruption": {
            "absolute_volatility_ratio": NaN,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": NaN,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": NaN,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 5,
            "mean_loss_difference": 0.0,
            "planned_items": 5,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "estimate": null,
            "matched_items": 3
          }
        },
        "cs-v7-arithmetic-semantic-crate-reconciliation": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.8134892168199606,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.7352941176470587,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.3074846997479609,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 6,
            "mean_loss_difference": -0.027777777777777846,
            "planned_items": 6,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.3972762620115435,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8187403993855605,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.19998821804272487,
            "majority_failure_difference": -0.33333333333333337,
            "matched_fraction": 1.0,
            "matched_items": 6,
            "mean_loss_difference": -0.05555555555555547,
            "planned_items": 6,
            "shared_condition": "A_shared"
          }
        },
        "cs-v7-arithmetic-semantic-shift-loss-recovery": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.8808303292720554,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.8463949843260192,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.16676914378646931,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 9,
            "mean_loss_difference": -0.037037037037037035,
            "planned_items": 9,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9113973673926883,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.9967741935483875,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.0032310205814460864,
            "majority_failure_difference": 0.125,
            "matched_fraction": 0.8888888888888888,
            "matched_items": 8,
            "mean_loss_difference": -0.02083333333333337,
            "planned_items": 9,
            "shared_condition": "A_shared"
          }
        },
        "cs-v7-code-fresh-layered-config-precedence": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.8674533114380004,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.1287128712871286,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.12107793155323592,
            "majority_failure_difference": 0.09999999999999998,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.08333333333333337,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.9595917942265422,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 2.1688888888888886,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.774215003648727,
            "majority_failure_difference": -0.30000000000000004,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.050000000000000044,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "cs-v7-code-fresh-versioned-transition-policy": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.084652289093281,
            "all_failure_difference": 0.1,
            "dependence_inflation_ratio": 1.4013840830449826,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.3374603789941069,
            "majority_failure_difference": 0.19999999999999996,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": 0.08333333333333337,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.7320508075688776,
            "all_failure_difference": 0.1111111111111111,
            "dependence_inflation_ratio": 3.25,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 1.1786549963416462,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": 1.1102230246251565e-16,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "cs-v7-constraint-fresh-multi-adjustment-band-policy": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.2428010522386699,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.6947194719471954,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.5275172238592442,
            "majority_failure_difference": -0.29999999999999993,
            "matched_fraction": 0.9090909090909091,
            "matched_items": 10,
            "mean_loss_difference": -0.08333333333333331,
            "planned_items": 11,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.2419230577438893,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.6587783818356248,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.5060814168935587,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 11,
            "mean_loss_difference": 0.015151515151515027,
            "planned_items": 11,
            "shared_condition": "A_shared"
          }
        },
        "cs-v7-constraint-fresh-ordered-policy-exceptions": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.299038105676658,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.6031250000000001,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.4719548493769974,
            "majority_failure_difference": 0.2222222222222222,
            "matched_fraction": 1.0,
            "matched_items": 9,
            "mean_loss_difference": 0.037037037037037035,
            "planned_items": 9,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.0,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.0,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.0,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 9,
            "mean_loss_difference": 0.0,
            "planned_items": 9,
            "shared_condition": "A_shared"
          }
        },
        "cs-v7-extraction-fresh-authoritative-audit-reduction": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 2.138089935299395,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 3.6770186335403725,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 1.3021022698995426,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.8,
            "matched_items": 8,
            "mean_loss_difference": -0.0625,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.7071067811865476,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.6071428571428572,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.49899116611898775,
            "majority_failure_difference": 0.11111111111111116,
            "matched_fraction": 0.9,
            "matched_items": 9,
            "mean_loss_difference": 0.0,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "cs-v7-extraction-fresh-cross-record-final-join": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 0.765092055676006,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.5681492109038737,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": -0.5653711995060434,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.016666666666666607,
            "planned_items": 10,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 0.9147320339189785,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 0.9023609443777513,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": -0.10274067889817347,
            "majority_failure_difference": -0.09999999999999998,
            "matched_fraction": 1.0,
            "matched_items": 10,
            "mean_loss_difference": -0.03333333333333327,
            "planned_items": 10,
            "shared_condition": "A_shared"
          }
        },
        "cs-v7-tool_call-conditional-incident-response-dag": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.914854215512676,
            "all_failure_difference": -0.25,
            "dependence_inflation_ratio": 1.5098039215686274,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.41197978912935806,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 8,
            "mean_loss_difference": -0.08333333333333326,
            "planned_items": 8,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": 1.0,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.0,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": 0.0,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 8,
            "mean_loss_difference": 0.0,
            "planned_items": 8,
            "shared_condition": "A_shared"
          }
        },
        "cs-v7-tool_call-multi-region-release-dag": {
          "evidence_corruption": {
            "absolute_volatility_ratio": 1.0,
            "all_failure_difference": 0.0,
            "dependence_inflation_ratio": 1.0,
            "independent_condition": "E_independent",
            "log_dependence_inflation_ratio": 0.0,
            "majority_failure_difference": 0.0,
            "matched_fraction": 1.0,
            "matched_items": 12,
            "mean_loss_difference": 1.1102230246251565e-16,
            "planned_items": 12,
            "shared_condition": "E_shared"
          },
          "instruction_anchor": {
            "absolute_volatility_ratio": NaN,
            "all_failure_difference": -0.18181818181818177,
            "dependence_inflation_ratio": NaN,
            "independent_condition": "A_independent",
            "log_dependence_inflation_ratio": NaN,
            "majority_failure_difference": 0.0,
            "matched_fraction": 0.9166666666666666,
            "matched_items": 11,
            "mean_loss_difference": -0.030303030303030387,
            "planned_items": 12,
            "shared_condition": "A_shared"
          }
        }
      }
    },
    "inference": "exploratory point estimates; no multiplicity-adjusted claims",
    "leave_one_template_family_out": {
      "cs-v7-arithmetic-lower-tier-batch-reconciliation": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 1.0093335966637451,
          "all_failure_difference": -0.010869565217391297,
          "dependence_inflation_ratio": 1.0103503879918665,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": 0.010297189494581983,
          "majority_failure_difference": 0.021739130434782594,
          "matched_fraction": 0.968421052631579,
          "matched_items": 92,
          "mean_loss_difference": -0.007246376811594235,
          "planned_items": 95,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.038103994396811,
          "all_failure_difference": -0.010989010989010978,
          "dependence_inflation_ratio": 1.042823231898663,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.041931681212884965,
          "majority_failure_difference": -0.04395604395604402,
          "matched_fraction": 0.9578947368421052,
          "matched_items": 91,
          "mean_loss_difference": -0.016483516483516536,
          "planned_items": 95,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-arithmetic-semantic-crate-reconciliation": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 1.0115420857800963,
          "all_failure_difference": -0.010989010989010978,
          "dependence_inflation_ratio": 1.0245792600594115,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": 0.024282050330401737,
          "majority_failure_difference": 0.02197802197802201,
          "matched_fraction": 0.9680851063829787,
          "matched_items": 91,
          "mean_loss_difference": -0.005494505494505586,
          "planned_items": 94,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.0226781899265596,
          "all_failure_difference": -0.011363636363636354,
          "dependence_inflation_ratio": 1.0235269740114121,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.02325448044844283,
          "majority_failure_difference": -0.022727272727272707,
          "matched_fraction": 0.9361702127659575,
          "matched_items": 88,
          "mean_loss_difference": -0.013257575757575912,
          "planned_items": 94,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-arithmetic-semantic-shift-loss-recovery": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 1.0007865114522652,
          "all_failure_difference": -0.011363636363636354,
          "dependence_inflation_ratio": 1.008065656465935,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": 0.008033302910831892,
          "majority_failure_difference": 0.022727272727272707,
          "matched_fraction": 0.967032967032967,
          "matched_items": 88,
          "mean_loss_difference": -0.0037878787878787845,
          "planned_items": 91,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.0393373957585774,
          "all_failure_difference": -0.011627906976744179,
          "dependence_inflation_ratio": 1.0521047483830255,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.05079268006498236,
          "majority_failure_difference": -0.05813953488372092,
          "matched_fraction": 0.945054945054945,
          "matched_items": 86,
          "mean_loss_difference": -0.015503875968992276,
          "planned_items": 91,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-code-fresh-layered-config-precedence": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 0.999060717672638,
          "all_failure_difference": -0.011494252873563232,
          "dependence_inflation_ratio": 0.9988412532394935,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": -0.001159418626599281,
          "majority_failure_difference": 0.011494252873563204,
          "matched_fraction": 0.9666666666666667,
          "matched_items": 87,
          "mean_loss_difference": -0.017241379310344862,
          "planned_items": 90,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.01430133678333,
          "all_failure_difference": -0.011904761904761918,
          "dependence_inflation_ratio": 1.019967567072816,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.0197708298032096,
          "majority_failure_difference": -0.011904761904761973,
          "matched_fraction": 0.9333333333333333,
          "matched_items": 84,
          "mean_loss_difference": -0.011904761904761751,
          "planned_items": 90,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-code-fresh-versioned-transition-policy": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 0.9960040687303093,
          "all_failure_difference": -0.022988505747126436,
          "dependence_inflation_ratio": 0.9929624110090782,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": -0.007062469622146706,
          "majority_failure_difference": 0.0,
          "matched_fraction": 0.9666666666666667,
          "matched_items": 87,
          "mean_loss_difference": -0.017241379310344862,
          "planned_items": 90,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.0198790857132933,
          "all_failure_difference": -0.023529411764705882,
          "dependence_inflation_ratio": 1.009940415454548,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.009891334513460767,
          "majority_failure_difference": -0.04705882352941182,
          "matched_fraction": 0.9444444444444444,
          "matched_items": 85,
          "mean_loss_difference": -0.01764705882352935,
          "planned_items": 90,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-constraint-fresh-multi-adjustment-band-policy": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 0.9884212764140778,
          "all_failure_difference": -0.011494252873563232,
          "dependence_inflation_ratio": 0.9668976111650118,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": -0.03366267210794093,
          "majority_failure_difference": 0.05747126436781613,
          "matched_fraction": 0.9775280898876404,
          "matched_items": 87,
          "mean_loss_difference": 0.0019157088122605526,
          "planned_items": 89,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.0190532869664612,
          "all_failure_difference": -0.01204819277108432,
          "dependence_inflation_ratio": 1.0105757656537853,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.010520233431587351,
          "majority_failure_difference": -0.048192771084337394,
          "matched_fraction": 0.9325842696629213,
          "matched_items": 83,
          "mean_loss_difference": -0.02008032128514059,
          "planned_items": 89,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-constraint-fresh-ordered-policy-exceptions": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 1.0187269722660044,
          "all_failure_difference": -0.011363636363636354,
          "dependence_inflation_ratio": 1.0214423171515177,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": 0.02121566491118983,
          "majority_failure_difference": 0.0,
          "matched_fraction": 0.967032967032967,
          "matched_items": 88,
          "mean_loss_difference": -0.011363636363636465,
          "planned_items": 91,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.0479045641350972,
          "all_failure_difference": -0.011764705882352955,
          "dependence_inflation_ratio": 1.0500001265217023,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.04879028466628405,
          "majority_failure_difference": -0.04705882352941171,
          "matched_fraction": 0.9340659340659341,
          "matched_items": 85,
          "mean_loss_difference": -0.017647058823529238,
          "planned_items": 91,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-extraction-fresh-authoritative-audit-reduction": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 1.003380672252923,
          "all_failure_difference": -0.011235955056179775,
          "dependence_inflation_ratio": 0.9980680259214614,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": -0.0019338427476601231,
          "majority_failure_difference": 0.022471910112359605,
          "matched_fraction": 0.9888888888888889,
          "matched_items": 89,
          "mean_loss_difference": -0.0018726591760299671,
          "planned_items": 90,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.0359512796696677,
          "all_failure_difference": -0.011764705882352955,
          "dependence_inflation_ratio": 1.0468134093273913,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.04575070141927451,
          "majority_failure_difference": -0.05882352941176472,
          "matched_fraction": 0.9444444444444444,
          "matched_items": 85,
          "mean_loss_difference": -0.01764705882352935,
          "planned_items": 90,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-extraction-fresh-cross-record-final-join": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 1.0078645158585064,
          "all_failure_difference": -0.011494252873563232,
          "dependence_inflation_ratio": 1.0092146638977655,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": 0.00917246789894685,
          "majority_failure_difference": 0.02298850574712641,
          "matched_fraction": 0.9666666666666667,
          "matched_items": 87,
          "mean_loss_difference": -0.005747126436781547,
          "planned_items": 90,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.0297683256777632,
          "all_failure_difference": -0.011904761904761918,
          "dependence_inflation_ratio": 1.0337527244061324,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.0331956028116696,
          "majority_failure_difference": -0.0357142857142857,
          "matched_fraction": 0.9333333333333333,
          "matched_items": 84,
          "mean_loss_difference": -0.01388888888888895,
          "planned_items": 90,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-tool_call-conditional-incident-response-dag": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 1.0357183103048113,
          "all_failure_difference": 0.011235955056179775,
          "dependence_inflation_ratio": 1.0585911452811811,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": 0.05693891585092077,
          "majority_failure_difference": 0.022471910112359605,
          "matched_fraction": 0.967391304347826,
          "matched_items": 89,
          "mean_loss_difference": 0.0,
          "planned_items": 92,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.0277537724664052,
          "all_failure_difference": -0.011627906976744193,
          "dependence_inflation_ratio": 1.0214225228044411,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.021196285903151688,
          "majority_failure_difference": -0.046511627906976716,
          "matched_fraction": 0.9347826086956522,
          "matched_items": 86,
          "mean_loss_difference": -0.01744186046511642,
          "planned_items": 92,
          "shared_condition": "A_shared"
        }
      },
      "cs-v7-tool_call-multi-region-release-dag": {
        "evidence_corruption": {
          "absolute_volatility_ratio": 1.0015396467931827,
          "all_failure_difference": -0.011764705882352941,
          "dependence_inflation_ratio": 1.0014532847737778,
          "independent_condition": "E_independent",
          "log_dependence_inflation_ratio": 0.0014522297774773188,
          "majority_failure_difference": 0.02352941176470591,
          "matched_fraction": 0.9659090909090909,
          "matched_items": 85,
          "mean_loss_difference": -0.007843137254901933,
          "planned_items": 88,
          "shared_condition": "E_shared"
        },
        "instruction_anchor": {
          "absolute_volatility_ratio": 1.0528340609394484,
          "all_failure_difference": 0.012048192771084335,
          "dependence_inflation_ratio": 1.0721587955359237,
          "independent_condition": "A_independent",
          "log_dependence_inflation_ratio": 0.06967418184252878,
          "majority_failure_difference": -0.04819277108433728,
          "matched_fraction": 0.9431818181818182,
          "matched_items": 83,
          "mean_loss_difference": -0.014056224899598346,
          "planned_items": 88,
          "shared_condition": "A_shared"
        }
      }
    },
    "schema_version": "cmfc.common-shock-subgroups.v1"
  },
  "verdict_outcomes": {
    "model_failure": 2014,
    "nan": 15,
    "pass": 1571
  }
}
