{
 "protocol": "heldout_v1",
 "generated": {
  "timestamp": "2026-11-04T00:51:23+00:00",
  "code_version": "760cf48-dirty",
  "host": {
   "platform": "macOS-27.0-arm64-arm-64bit",
   "python": "3.12.13",
   "machine": "arm64",
   "node": "MacBookPro-de-simon-pierre-2.local"
  }
 },
 "n_runs": 199,
 "n_runs_ok": 199,
 "n_judgments": 33,
 "hypotheses": {
  "H1_D_vs_B": {
   "clarity": {
    "n": 24,
    "mean_diff": -0.125,
    "ci95": [
     -0.312,
     0.031
    ],
    "wilcoxon_p": 0.17971249487899976,
    "cliffs_delta": -0.118,
    "holm_p": 0.7188,
    "significant": false
   },
   "argument_logic": {
    "n": 24,
    "mean_diff": 0.042,
    "ci95": [
     -0.156,
     0.219
    ],
    "wilcoxon_p": 0.5637028616507731,
    "cliffs_delta": 0.04,
    "holm_p": 1.0,
    "significant": false
   },
   "contribution_positioning": {
    "n": 24,
    "mean_diff": 0.0,
    "ci95": [
     -0.281,
     0.281
    ],
    "wilcoxon_p": 1.0,
    "cliffs_delta": 0.014,
    "holm_p": 1.0,
    "significant": false
   },
   "usefulness": {
    "n": 24,
    "mean_diff": 0.042,
    "ci95": [
     -0.281,
     0.406
    ],
    "wilcoxon_p": 0.7815112949987133,
    "cliffs_delta": 0.038,
    "holm_p": 1.0,
    "significant": false
   }
  },
  "H2_D_vs_C": {
   "clarity": {
    "n": 24,
    "mean_diff": -0.042,
    "ci95": [
     -0.094,
     0.0
    ],
    "wilcoxon_p": 0.31731050786291415,
    "cliffs_delta": -0.042,
    "holm_p": 0.9519,
    "significant": false
   },
   "argument_logic": {
    "n": 24,
    "mean_diff": -0.125,
    "ci95": [
     -0.219,
     0.031
    ],
    "wilcoxon_p": 0.17971249487899976,
    "cliffs_delta": -0.12,
    "holm_p": 0.7188,
    "significant": false
   },
   "contribution_positioning": {
    "n": 24,
    "mean_diff": 0.125,
    "ci95": [
     -0.031,
     0.219
    ],
    "wilcoxon_p": 0.31731050786291415,
    "cliffs_delta": 0.132,
    "holm_p": 0.6346,
    "significant": false
   },
   "usefulness": {
    "n": 24,
    "mean_diff": 0.083,
    "ci95": [
     -0.062,
     0.344
    ],
    "wilcoxon_p": 0.47950012218695337,
    "cliffs_delta": 0.069,
    "holm_p": 0.4795,
    "significant": false
   }
  },
  "H3_E_vs_D": {
   "methodological_rigor": {
    "n": 24,
    "mean_diff": 0.958,
    "ci95": [
     0.438,
     1.469
    ],
    "wilcoxon_p": 0.0028308913160744967,
    "cliffs_delta": 0.55,
    "holm_p": 0.0085,
    "significant": true
   },
   "fidelity": {
    "n": 24,
    "mean_diff": 0.458,
    "ci95": [
     0.031,
     0.688
    ],
    "wilcoxon_p": 0.011831666473446543,
    "cliffs_delta": 0.323,
    "holm_p": 0.0237,
    "significant": true
   },
   "usefulness": {
    "n": 24,
    "mean_diff": -0.333,
    "ci95": [
     -0.781,
     0.062
    ],
    "wilcoxon_p": 0.06462747314963886,
    "cliffs_delta": -0.243,
    "holm_p": 0.0646,
    "significant": false
   }
  },
  "E_vs_A": {
   "clarity": {
    "n": 24,
    "mean_diff": 0.042,
    "ci95": [
     -0.25,
     0.688
    ],
    "wilcoxon_p": 0.7388826803635273,
    "cliffs_delta": -0.111,
    "holm_p": 0.7389,
    "significant": false
   },
   "argument_logic": {
    "n": 24,
    "mean_diff": 0.458,
    "ci95": [
     0.094,
     1.313
    ],
    "wilcoxon_p": 0.11558942362652867,
    "cliffs_delta": 0.245,
    "holm_p": 0.4624,
    "significant": false
   },
   "contribution_positioning": {
    "n": 24,
    "mean_diff": -0.083,
    "ci95": [
     -0.375,
     0.719
    ],
    "wilcoxon_p": 0.4828996977801915,
    "cliffs_delta": -0.137,
    "holm_p": 0.9658,
    "significant": false
   },
   "methodological_rigor": {
    "n": 24,
    "mean_diff": 1.333,
    "ci95": [
     0.844,
     2.282
    ],
    "wilcoxon_p": 0.0003879158284341555,
    "cliffs_delta": 0.651,
    "holm_p": 0.0023,
    "significant": true
   },
   "fidelity": {
    "n": 24,
    "mean_diff": 0.375,
    "ci95": [
     -0.031,
     1.125
    ],
    "wilcoxon_p": 0.0577795711235972,
    "cliffs_delta": 0.22,
    "holm_p": 0.2889,
    "significant": false
   },
   "usefulness": {
    "n": 24,
    "mean_diff": -0.167,
    "ci95": [
     -0.562,
     0.438
    ],
    "wilcoxon_p": 0.3142753351370735,
    "cliffs_delta": -0.215,
    "holm_p": 0.9428,
    "significant": false
   }
  },
  "E_vs_B": {
   "clarity": {
    "n": 24,
    "mean_diff": -0.208,
    "ci95": [
     -0.469,
     -0.062
    ],
    "wilcoxon_p": 0.13166801602281422,
    "cliffs_delta": -0.193,
    "holm_p": 0.1317,
    "significant": false
   },
   "argument_logic": {
    "n": 24,
    "mean_diff": 0.333,
    "ci95": [
     0.094,
     0.688
    ],
    "wilcoxon_p": 0.08288063685529753,
    "cliffs_delta": 0.271,
    "holm_p": 0.2486,
    "significant": false
   },
   "contribution_positioning": {
    "n": 24,
    "mean_diff": -0.417,
    "ci95": [
     -0.75,
     0.094
    ],
    "wilcoxon_p": 0.07260252961103252,
    "cliffs_delta": -0.29,
    "holm_p": 0.2904,
    "significant": false
   },
   "methodological_rigor": {
    "n": 24,
    "mean_diff": 1.083,
    "ci95": [
     0.562,
     1.812
    ],
    "wilcoxon_p": 0.0014260915957594058,
    "cliffs_delta": 0.545,
    "holm_p": 0.0086,
    "significant": true
   },
   "fidelity": {
    "n": 24,
    "mean_diff": 0.375,
    "ci95": [
     0.156,
     0.594
    ],
    "wilcoxon_p": 0.012554918596966537,
    "cliffs_delta": 0.299,
    "holm_p": 0.0628,
    "significant": false
   },
   "usefulness": {
    "n": 24,
    "mean_diff": -0.292,
    "ci95": [
     -0.657,
     0.031
    ],
    "wilcoxon_p": 0.11709281778662677,
    "cliffs_delta": -0.207,
    "holm_p": 0.2342,
    "significant": false
   }
  }
 },
 "ablations": {
  "E_no_retrieval": {
   "clarity": {
    "n": 16,
    "mean_diff": 0.0,
    "ci95": [
     -0.25,
     0.312
    ],
    "wilcoxon_p": 1.0,
    "cliffs_delta": -0.008
   },
   "argument_logic": {
    "n": 16,
    "mean_diff": 0.188,
    "ci95": [
     -0.125,
     0.5
    ],
    "wilcoxon_p": 0.25683925795785656,
    "cliffs_delta": 0.176
   },
   "contribution_positioning": {
    "n": 16,
    "mean_diff": -0.25,
    "ci95": [
     -0.562,
     0.062
    ],
    "wilcoxon_p": 0.15729920705028505,
    "cliffs_delta": -0.125
   },
   "methodological_rigor": {
    "n": 16,
    "mean_diff": -0.062,
    "ci95": [
     -0.625,
     0.438
    ],
    "wilcoxon_p": 0.9416269114492879,
    "cliffs_delta": -0.02
   },
   "fidelity": {
    "n": 16,
    "mean_diff": 0.062,
    "ci95": [
     -0.188,
     0.375
    ],
    "wilcoxon_p": 0.6547208460185769,
    "cliffs_delta": 0.082
   },
   "usefulness": {
    "n": 16,
    "mean_diff": -0.125,
    "ci95": [
     -0.562,
     0.312
    ],
    "wilcoxon_p": 0.5574801348628696,
    "cliffs_delta": -0.047
   }
  },
  "E_no_profile": {
   "clarity": {
    "n": 16,
    "mean_diff": -0.062,
    "ci95": [
     -0.25,
     0.125
    ],
    "wilcoxon_p": 0.5637028616507731,
    "cliffs_delta": -0.066
   },
   "argument_logic": {
    "n": 16,
    "mean_diff": 0.0,
    "ci95": [
     -0.188,
     0.188
    ],
    "wilcoxon_p": 1.0,
    "cliffs_delta": 0.0
   },
   "contribution_positioning": {
    "n": 16,
    "mean_diff": -0.188,
    "ci95": [
     -0.375,
     0.0
    ],
    "wilcoxon_p": 0.08326451666355039,
    "cliffs_delta": -0.113
   },
   "methodological_rigor": {
    "n": 16,
    "mean_diff": -0.25,
    "ci95": [
     -0.5,
     0.0
    ],
    "wilcoxon_p": 0.10247043485974945,
    "cliffs_delta": -0.148
   },
   "fidelity": {
    "n": 16,
    "mean_diff": 0.0,
    "ci95": [
     -0.25,
     0.25
    ],
    "wilcoxon_p": 1.0,
    "cliffs_delta": -0.043
   },
   "usefulness": {
    "n": 16,
    "mean_diff": -0.25,
    "ci95": [
     -0.438,
     -0.062
    ],
    "wilcoxon_p": 0.045500263896358396,
    "cliffs_delta": -0.176
   }
  },
  "E_no_referee": {
   "clarity": {
    "n": 16,
    "mean_diff": -0.125,
    "ci95": [
     -0.375,
     0.125
    ],
    "wilcoxon_p": 0.31731050786291415,
    "cliffs_delta": -0.125
   },
   "argument_logic": {
    "n": 16,
    "mean_diff": 0.5,
    "ci95": [
     0.062,
     0.938
    ],
    "wilcoxon_p": 0.045579154787925666,
    "cliffs_delta": 0.422
   },
   "contribution_positioning": {
    "n": 16,
    "mean_diff": 0.062,
    "ci95": [
     -0.375,
     0.438
    ],
    "wilcoxon_p": 0.763024600552995,
    "cliffs_delta": 0.094
   },
   "methodological_rigor": {
    "n": 16,
    "mean_diff": 1.25,
    "ci95": [
     0.625,
     1.875
    ],
    "wilcoxon_p": 0.003243891789291054,
    "cliffs_delta": 0.625
   },
   "fidelity": {
    "n": 16,
    "mean_diff": 0.562,
    "ci95": [
     0.125,
     1.125
    ],
    "wilcoxon_p": 0.04734500208525526,
    "cliffs_delta": 0.371
   },
   "usefulness": {
    "n": 16,
    "mean_diff": 0.125,
    "ci95": [
     -0.375,
     0.562
    ],
    "wilcoxon_p": 0.5929800980174267,
    "cliffs_delta": 0.156
   }
  },
  "E_no_factcheck": {
   "clarity": {
    "n": 16,
    "mean_diff": -0.125,
    "ci95": [
     -0.375,
     0.125
    ],
    "wilcoxon_p": 0.31731050786291415,
    "cliffs_delta": -0.117
   },
   "argument_logic": {
    "n": 16,
    "mean_diff": 0.062,
    "ci95": [
     -0.314,
     0.438
    ],
    "wilcoxon_p": 0.7388826803635273,
    "cliffs_delta": 0.059
   },
   "contribution_positioning": {
    "n": 16,
    "mean_diff": 0.0,
    "ci95": [
     -0.312,
     0.312
    ],
    "wilcoxon_p": 1.0,
    "cliffs_delta": 0.086
   },
   "methodological_rigor": {
    "n": 16,
    "mean_diff": -0.062,
    "ci95": [
     -0.562,
     0.375
    ],
    "wilcoxon_p": 1.0,
    "cliffs_delta": 0.004
   },
   "fidelity": {
    "n": 16,
    "mean_diff": -0.062,
    "ci95": [
     -0.375,
     0.25
    ],
    "wilcoxon_p": 0.7054569861112734,
    "cliffs_delta": -0.062
   },
   "usefulness": {
    "n": 16,
    "mean_diff": -0.188,
    "ci95": [
     -0.5,
     0.125
    ],
    "wilcoxon_p": 0.25683925795785656,
    "cliffs_delta": -0.098
   }
  }
 },
 "judge_means": {
  "A_baseline": {
   "clarity": 5.79,
   "argument_logic": 5.58,
   "contribution_positioning": 5.0,
   "methodological_rigor": 4.79,
   "fidelity": 5.33,
   "usefulness": 5.58,
   "n_invented_facts": 1.38,
   "n": 24
  },
  "B_strong_prompt": {
   "clarity": 6.04,
   "argument_logic": 5.71,
   "contribution_positioning": 5.33,
   "methodological_rigor": 5.04,
   "fidelity": 5.33,
   "usefulness": 5.71,
   "n_invented_facts": 1.62,
   "n": 24
  },
  "C_conventional_rag": {
   "clarity": 5.96,
   "argument_logic": 5.88,
   "contribution_positioning": 5.21,
   "methodological_rigor": 5.21,
   "fidelity": 5.25,
   "usefulness": 5.67,
   "n_invented_facts": 1.71,
   "n": 24
  },
  "D_profiles_retrieval": {
   "clarity": 5.92,
   "argument_logic": 5.75,
   "contribution_positioning": 5.33,
   "methodological_rigor": 5.17,
   "fidelity": 5.25,
   "usefulness": 5.75,
   "n_invented_facts": 1.88,
   "n": 24
  },
  "E_full_engine": {
   "clarity": 5.83,
   "argument_logic": 6.04,
   "contribution_positioning": 4.92,
   "methodological_rigor": 6.12,
   "fidelity": 5.71,
   "usefulness": 5.42,
   "n_invented_facts": 0.88,
   "n": 24
  },
  "E_no_factcheck": {
   "clarity": 6.0,
   "argument_logic": 6.06,
   "contribution_positioning": 4.94,
   "methodological_rigor": 6.19,
   "fidelity": 5.75,
   "usefulness": 5.62,
   "n_invented_facts": 0.88,
   "n": 16
  },
  "E_no_profile": {
   "clarity": 5.94,
   "argument_logic": 6.12,
   "contribution_positioning": 5.12,
   "methodological_rigor": 6.38,
   "fidelity": 5.69,
   "usefulness": 5.69,
   "n_invented_facts": 1.0,
   "n": 16
  },
  "E_no_referee": {
   "clarity": 6.0,
   "argument_logic": 5.62,
   "contribution_positioning": 4.88,
   "methodological_rigor": 4.88,
   "fidelity": 5.12,
   "usefulness": 5.31,
   "n_invented_facts": 1.81,
   "n": 16
  },
  "E_no_retrieval": {
   "clarity": 5.88,
   "argument_logic": 5.94,
   "contribution_positioning": 5.19,
   "methodological_rigor": 6.19,
   "fidelity": 5.62,
   "usefulness": 5.56,
   "n_invented_facts": 1.5,
   "n": 16
  }
 },
 "task_means": {
  "('draft_introduction', 'A_baseline')": {
   "clarity": 6.25,
   "argument_logic": 6.25,
   "contribution_positioning": 5.5,
   "methodological_rigor": 5.25,
   "fidelity": 5.5,
   "usefulness": 6.0
  },
  "('draft_introduction', 'B_strong_prompt')": {
   "clarity": 6.25,
   "argument_logic": 6.25,
   "contribution_positioning": 6.25,
   "methodological_rigor": 5.25,
   "fidelity": 5.75,
   "usefulness": 6.0
  },
  "('draft_introduction', 'C_conventional_rag')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.25,
   "methodological_rigor": 5.5,
   "fidelity": 5.25,
   "usefulness": 6.0
  },
  "('draft_introduction', 'D_profiles_retrieval')": {
   "clarity": 5.75,
   "argument_logic": 6.25,
   "contribution_positioning": 5.5,
   "methodological_rigor": 5.0,
   "fidelity": 4.75,
   "usefulness": 5.75
  },
  "('draft_introduction', 'E_full_engine')": {
   "clarity": 6.0,
   "argument_logic": 6.25,
   "contribution_positioning": 4.25,
   "methodological_rigor": 6.0,
   "fidelity": 6.0,
   "usefulness": 5.0
  },
  "('draft_introduction', 'E_no_factcheck')": {
   "clarity": 6.0,
   "argument_logic": 6.5,
   "contribution_positioning": 4.0,
   "methodological_rigor": 6.5,
   "fidelity": 6.0,
   "usefulness": 5.5
  },
  "('draft_introduction', 'E_no_profile')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 4.0,
   "methodological_rigor": 6.0,
   "fidelity": 6.0,
   "usefulness": 5.0
  },
  "('draft_introduction', 'E_no_referee')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 5.0,
   "fidelity": 4.0,
   "usefulness": 5.0
  },
  "('draft_introduction', 'E_no_retrieval')": {
   "clarity": 5.5,
   "argument_logic": 5.5,
   "contribution_positioning": 4.0,
   "methodological_rigor": 5.5,
   "fidelity": 6.0,
   "usefulness": 5.0
  },
  "('explain_specification', 'A_baseline')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 5.0,
   "fidelity": 5.5,
   "usefulness": 6.0
  },
  "('explain_specification', 'B_strong_prompt')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.5,
   "methodological_rigor": 5.5,
   "fidelity": 5.0,
   "usefulness": 6.0
  },
  "('explain_specification', 'C_conventional_rag')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.0,
   "fidelity": 5.0,
   "usefulness": 6.0
  },
  "('explain_specification', 'D_profiles_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.0,
   "fidelity": 5.5,
   "usefulness": 6.5
  },
  "('explain_specification', 'E_full_engine')": {
   "clarity": 5.5,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.0,
   "fidelity": 5.0,
   "usefulness": 5.0
  },
  "('explain_specification', 'E_no_factcheck')": {
   "clarity": 5.5,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.0,
   "fidelity": 5.5,
   "usefulness": 5.0
  },
  "('explain_specification', 'E_no_profile')": {
   "clarity": 5.5,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.0,
   "fidelity": 5.5,
   "usefulness": 5.0
  },
  "('explain_specification', 'E_no_referee')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 5.5,
   "fidelity": 5.0,
   "usefulness": 5.5
  },
  "('explain_specification', 'E_no_retrieval')": {
   "clarity": 5.5,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.5,
   "fidelity": 5.5,
   "usefulness": 5.5
  },
  "('flag_causal_claims', 'A_baseline')": {
   "clarity": 3.5,
   "argument_logic": 3.5,
   "contribution_positioning": 3.0,
   "methodological_rigor": 3.5,
   "fidelity": 3.5,
   "usefulness": 3.5
  },
  "('flag_causal_claims', 'B_strong_prompt')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 5.5,
   "fidelity": 5.5,
   "usefulness": 5.0
  },
  "('flag_causal_claims', 'C_conventional_rag')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 5.5,
   "fidelity": 6.0,
   "usefulness": 5.0
  },
  "('flag_causal_claims', 'D_profiles_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.0,
   "fidelity": 6.0,
   "usefulness": 5.5
  },
  "('flag_causal_claims', 'E_full_engine')": {
   "clarity": 5.5,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.5,
   "fidelity": 6.0,
   "usefulness": 4.5
  },
  "('flag_causal_claims', 'E_no_factcheck')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 7.0,
   "fidelity": 5.5,
   "usefulness": 5.0
  },
  "('flag_causal_claims', 'E_no_profile')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 7.0,
   "fidelity": 6.0,
   "usefulness": 5.5
  },
  "('flag_causal_claims', 'E_no_referee')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 4.5,
   "methodological_rigor": 5.5,
   "fidelity": 6.0,
   "usefulness": 4.5
  },
  "('flag_causal_claims', 'E_no_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.5,
   "methodological_rigor": 6.5,
   "fidelity": 5.5,
   "usefulness": 5.5
  },
  "('improve_abstract', 'A_baseline')": {
   "clarity": 6.0,
   "argument_logic": 5.75,
   "contribution_positioning": 5.5,
   "methodological_rigor": 4.0,
   "fidelity": 6.0,
   "usefulness": 6.0
  },
  "('improve_abstract', 'B_strong_prompt')": {
   "clarity": 6.25,
   "argument_logic": 5.25,
   "contribution_positioning": 5.0,
   "methodological_rigor": 4.0,
   "fidelity": 6.0,
   "usefulness": 5.75
  },
  "('improve_abstract', 'C_conventional_rag')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.25,
   "methodological_rigor": 4.5,
   "fidelity": 6.25,
   "usefulness": 5.75
  },
  "('improve_abstract', 'D_profiles_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 5.75,
   "methodological_rigor": 4.25,
   "fidelity": 6.0,
   "usefulness": 6.0
  },
  "('improve_abstract', 'E_full_engine')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.25,
   "methodological_rigor": 5.5,
   "fidelity": 6.0,
   "usefulness": 6.0
  },
  "('improve_abstract', 'E_no_factcheck')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.5,
   "methodological_rigor": 5.5,
   "fidelity": 5.5,
   "usefulness": 6.0
  },
  "('improve_abstract', 'E_no_profile')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.5,
   "methodological_rigor": 6.0,
   "fidelity": 6.0,
   "usefulness": 6.0
  },
  "('improve_abstract', 'E_no_referee')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 5.0,
   "methodological_rigor": 4.0,
   "fidelity": 6.0,
   "usefulness": 5.5
  },
  "('improve_abstract', 'E_no_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.5,
   "methodological_rigor": 6.0,
   "fidelity": 6.0,
   "usefulness": 6.0
  },
  "('improve_table', 'A_baseline')": {
   "clarity": 5.5,
   "argument_logic": 5.0,
   "contribution_positioning": 4.0,
   "methodological_rigor": 4.0,
   "fidelity": 5.0,
   "usefulness": 5.5
  },
  "('improve_table', 'B_strong_prompt')": {
   "clarity": 6.0,
   "argument_logic": 5.0,
   "contribution_positioning": 4.0,
   "methodological_rigor": 4.0,
   "fidelity": 5.0,
   "usefulness": 6.0
  },
  "('improve_table', 'C_conventional_rag')": {
   "clarity": 6.0,
   "argument_logic": 5.0,
   "contribution_positioning": 4.0,
   "methodological_rigor": 4.0,
   "fidelity": 4.5,
   "usefulness": 6.0
  },
  "('improve_table', 'D_profiles_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 5.0,
   "contribution_positioning": 4.0,
   "methodological_rigor": 4.5,
   "fidelity": 5.5,
   "usefulness": 6.0
  },
  "('improve_table', 'E_full_engine')": {
   "clarity": 5.5,
   "argument_logic": 6.0,
   "contribution_positioning": 4.5,
   "methodological_rigor": 6.0,
   "fidelity": 5.5,
   "usefulness": 5.5
  },
  "('improve_table', 'E_no_factcheck')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 4.0,
   "methodological_rigor": 5.5,
   "fidelity": 5.5,
   "usefulness": 6.0
  },
  "('improve_table', 'E_no_profile')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 5.5,
   "fidelity": 5.5,
   "usefulness": 6.0
  },
  "('improve_table', 'E_no_referee')": {
   "clarity": 6.0,
   "argument_logic": 5.0,
   "contribution_positioning": 4.0,
   "methodological_rigor": 4.5,
   "fidelity": 5.5,
   "usefulness": 6.0
  },
  "('improve_table', 'E_no_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.5,
   "fidelity": 5.0,
   "usefulness": 6.0
  },
  "('interpret_results', 'A_baseline')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 5.0,
   "methodological_rigor": 4.75,
   "fidelity": 5.0,
   "usefulness": 5.5
  },
  "('interpret_results', 'B_strong_prompt')": {
   "clarity": 5.75,
   "argument_logic": 5.5,
   "contribution_positioning": 5.0,
   "methodological_rigor": 4.75,
   "fidelity": 4.5,
   "usefulness": 5.25
  },
  "('interpret_results', 'C_conventional_rag')": {
   "clarity": 6.0,
   "argument_logic": 5.75,
   "contribution_positioning": 5.0,
   "methodological_rigor": 4.75,
   "fidelity": 4.25,
   "usefulness": 5.5
  },
  "('interpret_results', 'D_profiles_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 5.25,
   "methodological_rigor": 5.0,
   "fidelity": 4.25,
   "usefulness": 5.5
  },
  "('interpret_results', 'E_full_engine')": {
   "clarity": 5.75,
   "argument_logic": 6.0,
   "contribution_positioning": 4.75,
   "methodological_rigor": 6.75,
   "fidelity": 5.5,
   "usefulness": 5.5
  },
  "('interpret_results', 'E_no_factcheck')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.5,
   "fidelity": 6.0,
   "usefulness": 6.0
  },
  "('interpret_results', 'E_no_profile')": {
   "clarity": 6.0,
   "argument_logic": 6.5,
   "contribution_positioning": 5.0,
   "methodological_rigor": 7.0,
   "fidelity": 6.0,
   "usefulness": 6.0
  },
  "('interpret_results', 'E_no_referee')": {
   "clarity": 6.0,
   "argument_logic": 5.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 4.0,
   "fidelity": 4.0,
   "usefulness": 5.5
  },
  "('interpret_results', 'E_no_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.5,
   "fidelity": 5.5,
   "usefulness": 5.5
  },
  "('referee_report', 'A_baseline')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.25,
   "fidelity": 6.0,
   "usefulness": 5.75
  },
  "('referee_report', 'B_strong_prompt')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.75,
   "methodological_rigor": 6.75,
   "fidelity": 5.75,
   "usefulness": 6.0
  },
  "('referee_report', 'C_conventional_rag')": {
   "clarity": 6.0,
   "argument_logic": 6.25,
   "contribution_positioning": 5.75,
   "methodological_rigor": 6.5,
   "fidelity": 5.75,
   "usefulness": 5.75
  },
  "('referee_report', 'D_profiles_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.5,
   "methodological_rigor": 6.25,
   "fidelity": 5.75,
   "usefulness": 5.75
  },
  "('referee_report', 'E_full_engine')": {
   "clarity": 6.0,
   "argument_logic": 5.75,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.0,
   "fidelity": 6.0,
   "usefulness": 5.5
  },
  "('referee_report', 'E_no_factcheck')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.0,
   "fidelity": 6.0,
   "usefulness": 5.5
  },
  "('referee_report', 'E_no_profile')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.0,
   "methodological_rigor": 6.5,
   "fidelity": 5.5,
   "usefulness": 6.0
  },
  "('referee_report', 'E_no_referee')": {
   "clarity": 6.0,
   "argument_logic": 6.5,
   "contribution_positioning": 5.5,
   "methodological_rigor": 6.5,
   "fidelity": 5.5,
   "usefulness": 5.5
  },
  "('referee_report', 'E_no_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.5,
   "methodological_rigor": 6.0,
   "fidelity": 6.5,
   "usefulness": 5.5
  },
  "('rewrite_literature', 'A_baseline')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 6.0,
   "methodological_rigor": 4.5,
   "fidelity": 5.0,
   "usefulness": 5.5
  },
  "('rewrite_literature', 'B_strong_prompt')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 5.5,
   "methodological_rigor": 4.0,
   "fidelity": 4.5,
   "usefulness": 5.5
  },
  "('rewrite_literature', 'C_conventional_rag')": {
   "clarity": 5.5,
   "argument_logic": 5.5,
   "contribution_positioning": 6.0,
   "methodological_rigor": 4.5,
   "fidelity": 4.5,
   "usefulness": 5.0
  },
  "('rewrite_literature', 'D_profiles_retrieval')": {
   "clarity": 5.5,
   "argument_logic": 5.5,
   "contribution_positioning": 6.0,
   "methodological_rigor": 4.5,
   "fidelity": 4.5,
   "usefulness": 5.0
  },
  "('rewrite_literature', 'E_full_engine')": {
   "clarity": 6.0,
   "argument_logic": 6.5,
   "contribution_positioning": 6.0,
   "methodological_rigor": 6.5,
   "fidelity": 5.0,
   "usefulness": 6.0
  },
  "('rewrite_literature', 'E_no_factcheck')": {
   "clarity": 6.5,
   "argument_logic": 6.5,
   "contribution_positioning": 6.0,
   "methodological_rigor": 6.5,
   "fidelity": 6.0,
   "usefulness": 6.0
  },
  "('rewrite_literature', 'E_no_profile')": {
   "clarity": 6.0,
   "argument_logic": 7.0,
   "contribution_positioning": 6.5,
   "methodological_rigor": 7.0,
   "fidelity": 5.0,
   "usefulness": 6.0
  },
  "('rewrite_literature', 'E_no_referee')": {
   "clarity": 6.0,
   "argument_logic": 5.5,
   "contribution_positioning": 5.0,
   "methodological_rigor": 4.0,
   "fidelity": 5.0,
   "usefulness": 5.0
  },
  "('rewrite_literature', 'E_no_retrieval')": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 6.0,
   "methodological_rigor": 6.0,
   "fidelity": 5.0,
   "usefulness": 5.5
  }
 },
 "out_of_domain_means": {
  "A_baseline": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 5.67,
   "methodological_rigor": 5.67,
   "fidelity": 5.0,
   "usefulness": 5.33
  },
  "B_strong_prompt": {
   "clarity": 6.0,
   "argument_logic": 6.0,
   "contribution_positioning": 6.0,
   "methodological_rigor": 5.33,
   "fidelity": 4.67,
   "usefulness": 5.33
  },
  "C_conventional_rag": {
   "clarity": 6.0,
   "argument_logic": 5.67,
   "contribution_positioning": 5.67,
   "methodological_rigor": 5.0,
   "fidelity": 4.33,
   "usefulness": 5.33
  },
  "D_profiles_retrieval": {
   "clarity": 5.33,
   "argument_logic": 5.67,
   "contribution_positioning": 5.33,
   "methodological_rigor": 5.67,
   "fidelity": 4.67,
   "usefulness": 4.67
  },
  "E_full_engine": {
   "clarity": 5.67,
   "argument_logic": 6.0,
   "contribution_positioning": 5.33,
   "methodological_rigor": 5.67,
   "fidelity": 4.33,
   "usefulness": 5.0
  }
 },
 "judge_consistency": {
  "n_pairs": 50,
  "clarity": {
   "mean_abs_diff": 0.26,
   "spearman": 0.411
  },
  "argument_logic": {
   "mean_abs_diff": 0.3,
   "spearman": 0.582
  },
  "contribution_positioning": {
   "mean_abs_diff": 0.32,
   "spearman": 0.84
  },
  "methodological_rigor": {
   "mean_abs_diff": 0.34,
   "spearman": 0.804
  },
  "fidelity": {
   "mean_abs_diff": 0.3,
   "spearman": 0.747
  },
  "usefulness": {
   "mean_abs_diff": 0.4,
   "spearman": 0.531
  }
 },
 "secondary_means": {
  "A_baseline": {
   "preserved_rate": 0.053,
   "new_numbers": 1.0,
   "n_unverifiable_citations": 0.083,
   "n_unsupported_claims": 1.083,
   "max_overlap_share": 0.073,
   "todo_placeholders": 2.542,
   "n_words": 959.333,
   "cost_usd": 0.138,
   "latency_s": 45.879,
   "n": 24
  },
  "B_strong_prompt": {
   "preserved_rate": 0.057,
   "new_numbers": 1.042,
   "n_unverifiable_citations": 0.083,
   "n_unsupported_claims": 1.042,
   "max_overlap_share": 0.065,
   "todo_placeholders": 2.958,
   "n_words": 1027.417,
   "cost_usd": 0.139,
   "latency_s": 46.083,
   "n": 24
  },
  "C_conventional_rag": {
   "preserved_rate": 0.053,
   "new_numbers": 1.25,
   "n_unverifiable_citations": 0.083,
   "n_unsupported_claims": 1.625,
   "max_overlap_share": 0.072,
   "todo_placeholders": 2.167,
   "n_words": 1059.875,
   "cost_usd": 0.146,
   "latency_s": 47.446,
   "n": 24
  },
  "D_profiles_retrieval": {
   "preserved_rate": 0.055,
   "new_numbers": 1.167,
   "n_unverifiable_citations": 0.083,
   "n_unsupported_claims": 1.208,
   "max_overlap_share": 0.068,
   "todo_placeholders": 2.667,
   "n_words": 1051.875,
   "cost_usd": 0.154,
   "latency_s": 49.475,
   "n": 24
  },
  "E_full_engine": {
   "preserved_rate": 0.055,
   "new_numbers": 0.833,
   "n_unverifiable_citations": 0.0,
   "n_unsupported_claims": 0.625,
   "max_overlap_share": 0.059,
   "todo_placeholders": 3.125,
   "n_words": 994.667,
   "cost_usd": 0.316,
   "latency_s": 98.033,
   "n": 24
  },
  "E_no_factcheck": {
   "preserved_rate": 0.048,
   "new_numbers": 0.938,
   "n_unverifiable_citations": 0.062,
   "n_unsupported_claims": 0.75,
   "max_overlap_share": 0.057,
   "todo_placeholders": 2.562,
   "n_words": 1049.938,
   "cost_usd": 0.312,
   "latency_s": 95.925,
   "n": 16
  },
  "E_no_profile": {
   "preserved_rate": 0.054,
   "new_numbers": 0.688,
   "n_unverifiable_citations": 0.0,
   "n_unsupported_claims": 0.312,
   "max_overlap_share": 0.06,
   "todo_placeholders": 2.75,
   "n_words": 996.188,
   "cost_usd": 0.311,
   "latency_s": 97.706,
   "n": 16
  },
  "E_no_referee": {
   "preserved_rate": 0.044,
   "new_numbers": 0.875,
   "n_unverifiable_citations": 0.125,
   "n_unsupported_claims": 1.062,
   "max_overlap_share": 0.076,
   "todo_placeholders": 2.25,
   "n_words": 939.188,
   "cost_usd": 0.155,
   "latency_s": 48.856,
   "n": 16
  },
  "E_no_retrieval": {
   "preserved_rate": 0.042,
   "new_numbers": 0.625,
   "n_unverifiable_citations": 0.125,
   "n_unsupported_claims": 0.312,
   "max_overlap_share": 0.055,
   "todo_placeholders": 5.062,
   "n_words": 1074.375,
   "cost_usd": 0.308,
   "latency_s": 94.531,
   "n": 16
  }
 },
 "H4": {
  "E_vs_A": {
   "n_unverifiable_citations": {
    "n": 24,
    "mean_diff": -0.083,
    "ci95": [
     -0.25,
     0.0
    ],
    "wilcoxon_p": 0.15729920705028505,
    "cliffs_delta": -0.083
   },
   "n_unsupported_claims": {
    "n": 24,
    "mean_diff": -0.458,
    "ci95": [
     -0.906,
     0.25
    ],
    "wilcoxon_p": 0.1005626399172922,
    "cliffs_delta": -0.163
   },
   "new_numbers": {
    "n": 24,
    "mean_diff": -0.167,
    "ci95": [
     -0.375,
     0.125
    ],
    "wilcoxon_p": 0.15729920705028516,
    "cliffs_delta": -0.101
   }
  },
  "E_vs_B": {
   "n_unverifiable_citations": {
    "n": 24,
    "mean_diff": -0.083,
    "ci95": [
     -0.156,
     0.0
    ],
    "wilcoxon_p": 0.15729920705028505,
    "cliffs_delta": -0.083
   },
   "n_unsupported_claims": {
    "n": 24,
    "mean_diff": -0.417,
    "ci95": [
     -0.938,
     0.188
    ],
    "wilcoxon_p": 0.17298970944016634,
    "cliffs_delta": -0.181
   },
   "new_numbers": {
    "n": 24,
    "mean_diff": -0.208,
    "ci95": [
     -0.406,
     0.062
    ],
    "wilcoxon_p": 0.09558070454562939,
    "cliffs_delta": -0.141
   }
  },
  "E_vs_C": {
   "n_unverifiable_citations": {
    "n": 24,
    "mean_diff": -0.083,
    "ci95": [
     -0.312,
     0.0
    ],
    "wilcoxon_p": 0.15729920705028505,
    "cliffs_delta": -0.083
   },
   "n_unsupported_claims": {
    "n": 24,
    "mean_diff": -1.0,
    "ci95": [
     -1.594,
     -0.156
    ],
    "wilcoxon_p": 0.008554328993751146,
    "cliffs_delta": -0.366
   },
   "new_numbers": {
    "n": 24,
    "mean_diff": -0.417,
    "ci95": [
     -1.0,
     0.125
    ],
    "wilcoxon_p": 0.31731050786291415,
    "cliffs_delta": -0.082
   }
  }
 },
 "H5_overlap": {
  "preregistered_metric": {
   "max_overlap_share_overall": 0.4044,
   "share_outputs_below_0.02": 0.41847826086956524
  },
  "excluding_source_document": {
   "max_other_share": 0.0069,
   "share_outputs_below_0.02": 1.0,
   "mean_other_share": 0.0007923913043478261
  },
  "self_overlap_by_task": {
   "draft_introduction": 0.028,
   "explain_specification": 0.019,
   "flag_causal_claims": 0.073,
   "improve_abstract": 0.253,
   "improve_table": 0.02,
   "interpret_results": 0.018,
   "referee_report": 0.008,
   "rewrite_literature": 0.082
  },
  "self_overlap_by_condition": {
   "A_baseline": 0.073,
   "B_strong_prompt": 0.065,
   "C_conventional_rag": 0.072,
   "D_profiles_retrieval": 0.068,
   "E_full_engine": 0.059,
   "E_no_factcheck": 0.057,
   "E_no_profile": 0.06,
   "E_no_referee": 0.076,
   "E_no_retrieval": 0.055
  }
 }
}