{
  "analyses": {
    "gpt55_extension": {
      "direct_qwen_contrast": {
        "inference_role": "descriptive_unregistered_contrast",
        "judges": {
          "CerebrasGemma431b": {
            "ci90": [
              4.0,
              14.5
            ],
            "point": 9.5
          },
          "CerebrasGptOss120b": {
            "ci90": [
              -1.0,
              6.0
            ],
            "point": 2.5
          }
        },
        "registration_status": "post_hoc"
      },
      "registered_read": {
        "inference_role": "descriptive_extension",
        "judges": {
          "CerebrasGemma431b": {
            "gpt55_vs_gemma_baseline": {
              "ci90": [
                28.0,
                38.0
              ],
              "point": 33.5
            },
            "largest_arm": "gpt55_gemma_skilled_full",
            "largest_baseline_delta_on_gemma_column": true
          },
          "CerebrasGptOss120b": {
            "gpt55_vs_gemma_baseline": {
              "ci90": [
                18.5,
                25.0
              ],
              "point": 21.5
            },
            "largest_arm": "gpt55_gemma_skilled_full",
            "largest_baseline_delta_on_gemma_column": true
          }
        },
        "registration_status": "registered"
      }
    },
    "participant_transport": {
      "inference_role": "descriptive_sensitivity_no_frozen_rule",
      "judges": {
        "CerebrasGemma431b": {
          "arm_deltas_vs_baseline": {
            "gemma_echo_full": {
              "ci90": [
                10.0,
                17.0
              ],
              "point": 13.0
            },
            "gemma_echo_public": {
              "ci90": [
                10.5,
                18.0
              ],
              "point": 15.5
            },
            "gemma_skilled_full": {
              "ci90": [
                20.5,
                27.0
              ],
              "point": 24.0
            },
            "gemma_skilled_public": {
              "ci90": [
                14.5,
                23.0
              ],
              "point": 18.0
            }
          },
          "estimands": {
            "E1": {
              "ci90": [
                -0.5,
                9.5
              ],
              "point": 2.5
            },
            "E2": {
              "ci90": [
                6.0,
                14.0
              ],
              "point": 11.0
            },
            "E3": {
              "ci90": [
                -12.5,
                0.5
              ],
              "point": -8.5
            }
          },
          "minimum_mediated_arm_vs_baseline": {
            "arm": "gemma_echo_full",
            "ci90": [
              10.0,
              17.0
            ],
            "point": 13.0
          },
          "n_joint": 62
        },
        "CerebrasGptOss120b": {
          "arm_deltas_vs_baseline": {
            "gemma_echo_full": {
              "ci90": [
                9.0,
                12.5
              ],
              "point": 12.0
            },
            "gemma_echo_public": {
              "ci90": [
                10.5,
                15.0
              ],
              "point": 14.0
            },
            "gemma_skilled_full": {
              "ci90": [
                16.5,
                22.0
              ],
              "point": 19.0
            },
            "gemma_skilled_public": {
              "ci90": [
                12.5,
                19.0
              ],
              "point": 16.5
            }
          },
          "estimands": {
            "E1": {
              "ci90": [
                -1.0,
                6.5
              ],
              "point": 2.5
            },
            "E2": {
              "ci90": [
                5.5,
                11.0
              ],
              "point": 7.0
            },
            "E3": {
              "ci90": [
                -10.0,
                -0.5
              ],
              "point": -4.5
            }
          },
          "minimum_mediated_arm_vs_baseline": {
            "arm": "gemma_echo_full",
            "ci90": [
              9.0,
              12.5
            ],
            "point": 12.0
          },
          "n_joint": 62
        }
      },
      "registration_status": "registered"
    },
    "stheno_public_context_skill": {
      "judges": {
        "CerebrasGemma431b": {
          "ci90": [
            11.0,
            21.75
          ],
          "point": 18.75
        },
        "CerebrasGptOss120b": {
          "ci90": [
            4.5,
            11.5
          ],
          "point": 8.25
        }
      },
      "registration_status": "registered",
      "replicate_reduction": "per_scenario_arithmetic_mean"
    }
  },
  "artifact_date": "2026-07-28",
  "artifact_id": "mediationbench-crossed-readout-2026-07-28",
  "bootstrap": {
    "confidence_level": 0.9,
    "method": "joint_scenario_resampling_percentile",
    "samples": 10000,
    "seed": 20260728
  },
  "claim_bindings": {
    "CFBaseMinGemmaJ": {
      "registration_status": "registered",
      "source": "analyses.participant_transport.judges.CerebrasGemma431b.minimum_mediated_arm_vs_baseline"
    },
    "CFBaseMinOssJ": {
      "registration_status": "registered",
      "source": "analyses.participant_transport.judges.CerebrasGptOss120b.minimum_mediated_arm_vs_baseline"
    },
    "CFEOneGemmaJ": {
      "registration_status": "registered",
      "source": "analyses.participant_transport.judges.CerebrasGemma431b.estimands.E1"
    },
    "CFEOneOssJ": {
      "registration_status": "registered",
      "source": "analyses.participant_transport.judges.CerebrasGptOss120b.estimands.E1"
    },
    "CFEThreeGemmaJ": {
      "registration_status": "registered",
      "source": "analyses.participant_transport.judges.CerebrasGemma431b.estimands.E3"
    },
    "CFEThreeOssJ": {
      "registration_status": "registered",
      "source": "analyses.participant_transport.judges.CerebrasGptOss120b.estimands.E3"
    },
    "CFETwoGemmaJ": {
      "registration_status": "registered",
      "source": "analyses.participant_transport.judges.CerebrasGemma431b.estimands.E2"
    },
    "CFETwoOssJ": {
      "registration_status": "registered",
      "source": "analyses.participant_transport.judges.CerebrasGptOss120b.estimands.E2"
    },
    "CFSthenoBlindGemmaJ": {
      "registration_status": "registered",
      "source": "analyses.stheno_public_context_skill.judges.CerebrasGemma431b"
    },
    "CFSthenoBlindOssJ": {
      "registration_status": "registered",
      "source": "analyses.stheno_public_context_skill.judges.CerebrasGptOss120b"
    },
    "JudgeSweepGemma": {
      "registration_status": "design_metadata",
      "source": "judges.CerebrasGemma431b.display_name"
    },
    "JudgeSweepOss": {
      "registration_status": "design_metadata",
      "source": "judges.CerebrasGptOss120b.display_name"
    }
  },
  "coverage": {
    "all_pairs_same_scenario_set": true,
    "complete": true,
    "expected_pair_count": 36,
    "expected_scenarios_per_pair": 62,
    "pairs": [
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "kimi_echo_full",
        "run_id": "83a48744-cb7b-4794-a5a2-93c7c930603d"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "kimi_echo_full",
        "run_id": "83a48744-cb7b-4794-a5a2-93c7c930603d"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "kimi_echo_public",
        "run_id": "1b711261-7c3d-47c3-866b-2f7dc415a490"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "kimi_echo_public",
        "run_id": "1b711261-7c3d-47c3-866b-2f7dc415a490"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_echo_full_b",
        "run_id": "83752dc7-8edf-4a3d-8a5e-aceeb6e29cb0"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_echo_full_b",
        "run_id": "83752dc7-8edf-4a3d-8a5e-aceeb6e29cb0"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_echo_full_a",
        "run_id": "df31ef89-7280-4ce6-b625-db84c31687f0"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_echo_full_a",
        "run_id": "df31ef89-7280-4ce6-b625-db84c31687f0"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_echo_public_b",
        "run_id": "f0b1d261-9f3d-40b6-95df-f1b8b6245149"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_echo_public_b",
        "run_id": "f0b1d261-9f3d-40b6-95df-f1b8b6245149"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_echo_public_a",
        "run_id": "77e6b874-3163-4955-ba13-aa6d0f4031d4"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_echo_public_a",
        "run_id": "77e6b874-3163-4955-ba13-aa6d0f4031d4"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gptoss_echo_full",
        "run_id": "cd46badc-69cc-4bec-a22a-60604567c7fb"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gptoss_echo_full",
        "run_id": "cd46badc-69cc-4bec-a22a-60604567c7fb"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gptoss_echo_public",
        "run_id": "5f378266-2439-40ac-aedc-24aceb79ef3e"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gptoss_echo_public",
        "run_id": "5f378266-2439-40ac-aedc-24aceb79ef3e"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gptoss_skilled_public",
        "run_id": "d82f5d2a-1fd3-4c5b-b011-a28661c7e4d5"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gptoss_skilled_public",
        "run_id": "d82f5d2a-1fd3-4c5b-b011-a28661c7e4d5"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "kimi_skilled_public",
        "run_id": "726730c8-6791-43ad-ad3b-02e50bd7465c"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "kimi_skilled_public",
        "run_id": "726730c8-6791-43ad-ad3b-02e50bd7465c"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_skilled_public_b",
        "run_id": "0e3e0f47-e0cf-4a82-be0e-42bd2432e8d2"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_skilled_public_b",
        "run_id": "0e3e0f47-e0cf-4a82-be0e-42bd2432e8d2"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_skilled_public_a",
        "run_id": "986c4f8c-bbf4-4c11-9f09-11b3c7a39d43"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "qwen_skilled_public_a",
        "run_id": "986c4f8c-bbf4-4c11-9f09-11b3c7a39d43"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_echo_full",
        "run_id": "c4f0de97-f99a-4775-a3ff-eaa7a80051d7"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_echo_full",
        "run_id": "c4f0de97-f99a-4775-a3ff-eaa7a80051d7"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_echo_public",
        "run_id": "0f77a820-61f0-4b75-9b7c-3fd51af35589"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_echo_public",
        "run_id": "0f77a820-61f0-4b75-9b7c-3fd51af35589"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_skilled_full",
        "run_id": "29ee60cc-7c87-4cf0-acb9-37fca9c014db"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_skilled_full",
        "run_id": "29ee60cc-7c87-4cf0-acb9-37fca9c014db"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_skilled_public",
        "run_id": "1650e2f0-9a95-4422-9aa9-97397c1114d9"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_skilled_public",
        "run_id": "1650e2f0-9a95-4422-9aa9-97397c1114d9"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_base",
        "run_id": "d7275905-c038-4ce6-bf58-66efaa6e399d"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gemma_base",
        "run_id": "d7275905-c038-4ce6-bf58-66efaa6e399d"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGemma431b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gpt55_gemma_skilled_full",
        "run_id": "1e91a777-465a-498a-bfbb-27bf9e2b09b4"
      },
      {
        "complete": true,
        "judge_id": "CerebrasGptOss120b",
        "n_expected": 62,
        "n_scored": 62,
        "run_alias": "gpt55_gemma_skilled_full",
        "run_id": "1e91a777-465a-498a-bfbb-27bf9e2b09b4"
      }
    ]
  },
  "estimator": {
    "contrast_orientation": "left_minus_right",
    "e1_orientation": "skilled_public_minus_echo_public",
    "e2_orientation": "skilled_full_minus_echo_full",
    "e3_orientation": "E1_minus_E2",
    "id": "difference_of_arm_medians",
    "joint_set_rule": "complete_case_across_all_registered_run_judge_pairs",
    "replicate_reduction": "per_scenario_arithmetic_mean"
  },
  "inputs": [
    {
      "bytes": 184045,
      "id": "crossed_input_001",
      "kind": "private_crossed_scores",
      "row_count": 2232,
      "sha256": "64c9c2ca13e0f5024e6bb49f8b3a06d0a0df8a42d58a762821bb7ba5e5063f0c"
    },
    {
      "bytes": 35068,
      "id": "operational_aggregate",
      "kind": "aggregate_claims_candidate",
      "sha256": "c1c53164a1441186e7369b8a1a3592a3dcf55021510322a68f0ef1d0c7e266b9"
    },
    {
      "bytes": 28353,
      "id": "operational_claims_generator",
      "kind": "generator_source",
      "sha256": "fc7b0b63f205e02359e93a5cf17fece90b36c8d79af1ad7f72a46304284b86f0"
    },
    {
      "bytes": 37497,
      "id": "crossed_readout_generator",
      "kind": "generator_source",
      "sha256": "603fbb7714241d015500d76ec2455001b9c94b5d0b11cc610556156054b7245e"
    }
  ],
  "judges": {
    "CerebrasGemma431b": {
      "display_name": "Gemma-4-31B",
      "registration_scope": "registered_cross_family_sweep"
    },
    "CerebrasGptOss120b": {
      "display_name": "gpt-oss-120b",
      "registration_scope": "registered_cross_family_sweep"
    }
  },
  "operational_reference": {
    "bootstrap": {
      "confidence_level": 0.9,
      "method": "joint_scenario_resampling_percentile",
      "samples": 10000,
      "seed": 20260706
    },
    "contrast_orientation": "left_minus_right",
    "e3_orientation": "E1_minus_E2",
    "estimator_id": "difference_of_arm_medians",
    "participant_transport": {
      "baseline_deltas": {
        "gemma/echo_blind": {
          "ci90": [
            14.15,
            25.06
          ],
          "delta_median": 18.63,
          "n_joint": 62
        },
        "gemma/echo_full": {
          "ci90": [
            14.35,
            22.22
          ],
          "delta_median": 17.9,
          "n_joint": 62
        },
        "gemma/skilled_blind": {
          "ci90": [
            16.99,
            30.71
          ],
          "delta_median": 24.33,
          "n_joint": 62
        },
        "gemma/skilled_full": {
          "ci90": [
            23.88,
            34.72
          ],
          "delta_median": 29.29,
          "n_joint": 62
        }
      },
      "estimands": {
        "E1": {
          "ci90": [
            -3.79,
            12.9
          ],
          "point": 5.7
        },
        "E2": {
          "ci90": [
            5.77,
            16.92
          ],
          "point": 11.39
        },
        "E3": {
          "ci90": [
            -15.95,
            2.3
          ],
          "point": -5.68
        }
      },
      "n_joint": 62
    },
    "source_input_id": "operational_aggregate",
    "source_sha256": "c1c53164a1441186e7369b8a1a3592a3dcf55021510322a68f0ef1d0c7e266b9"
  },
  "privacy": {
    "classification": "aggregate_only",
    "contains_fixtures": false,
    "contains_prompts": false,
    "contains_scenario_rows": false,
    "contains_transcripts": false
  },
  "provenance_limitations": [
    "historical_judge_prompt_version_and_text_hash_not_retained",
    "historical_rejudge_job_and_snapshot_ids_not_retained",
    "provider_response_model_revision_not_retained"
  ],
  "release_gate": {
    "blocking_conditions": [],
    "optional_gpt55_operational_score": "omitted_not_a_release_dependency",
    "review": "separate_internal_adversarial_artifact_and_code_review_2026-07-28",
    "scope": "aggregate fixed-transcript crossed-scorer claims only",
    "state": "passed"
  },
  "runs": {
    "gemma_base": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "d7275905-c038-4ce6-bf58-66efaa6e399d"
    },
    "gemma_echo_full": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "c4f0de97-f99a-4775-a3ff-eaa7a80051d7"
    },
    "gemma_echo_public": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "0f77a820-61f0-4b75-9b7c-3fd51af35589"
    },
    "gemma_skilled_full": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "29ee60cc-7c87-4cf0-acb9-37fca9c014db"
    },
    "gemma_skilled_public": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "1650e2f0-9a95-4422-9aa9-97397c1114d9"
    },
    "gpt55_gemma_skilled_full": {
      "display_label": "Skilled@GPT-5.5",
      "registration_scope": "registered_gpt55_extension",
      "run_id": "1e91a777-465a-498a-bfbb-27bf9e2b09b4"
    },
    "gptoss_echo_full": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "cd46badc-69cc-4bec-a22a-60604567c7fb"
    },
    "gptoss_echo_public": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "5f378266-2439-40ac-aedc-24aceb79ef3e"
    },
    "gptoss_skilled_public": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "d82f5d2a-1fd3-4c5b-b011-a28661c7e4d5"
    },
    "kimi_echo_full": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "83a48744-cb7b-4794-a5a2-93c7c930603d"
    },
    "kimi_echo_public": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "1b711261-7c3d-47c3-866b-2f7dc415a490"
    },
    "kimi_skilled_public": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "726730c8-6791-43ad-ad3b-02e50bd7465c"
    },
    "qwen_echo_full_a": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "df31ef89-7280-4ce6-b625-db84c31687f0"
    },
    "qwen_echo_full_b": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "83752dc7-8edf-4a3d-8a5e-aceeb6e29cb0"
    },
    "qwen_echo_public_a": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "77e6b874-3163-4955-ba13-aa6d0f4031d4"
    },
    "qwen_echo_public_b": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "f0b1d261-9f3d-40b6-95df-f1b8b6245149"
    },
    "qwen_skilled_public_a": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "986c4f8c-bbf4-4c11-9f09-11b3c7a39d43"
    },
    "qwen_skilled_public_b": {
      "registration_scope": "registered_cross_family_sweep",
      "run_id": "0e3e0f47-e0cf-4a82-be0e-42bd2432e8d2"
    }
  },
  "schema_version": "mediationbench.crossed_readout.v1",
  "status": "release_ready"
}
