1{2 "generated_at": "2026-10-28T08:49:55+00:00",3 "script": "scripts/20_evaluate.py",4 "runs": [5 "baselines",6 "final_E_single_llm_mixed_random_mid",7 "final_F_independent_mixed_random_mid",8 "final_G_debate_r2_mixed_random_mid",9 "final_H_debate_skeptic_r2_mixed_random_mid",10 "final_I_placebo_r2_mixed_random_mid",11 "final_J_cost_matched_mixed_random_mid",12 "final_small_E_single_llm_mixed_random_small",13 "final_small_F_independent_mixed_random_small",14 "final_small_G_debate_r2_mixed_random_small",15 "holdout_E_single_llm_mixed_random_mid",16 "holdout_F_independent_mixed_random_mid",17 "holdout_G_debate_r2_mixed_random_mid",18 "prospective_E_single_llm_mixed_random_mid",19 "prospective_F_independent_mixed_random_mid",20 "prospective_G_debate_r2_mixed_random_mid",21 "prospective_baselines",22 "validation_E_single_llm_all_anthropic_mid",23 "validation_E_single_llm_all_openai_mid",24 "validation_E_single_llm_mixed_random_mid",25 "validation_F_independent_all_anthropic_mid",26 "validation_F_independent_all_openai_mid",27 "validation_F_independent_mixed_random_mid",28 "validation_G_debate_r1_mixed_random_mid",29 "validation_G_debate_r2_all_anthropic_mid",30 "validation_G_debate_r2_all_openai_mid",31 "validation_G_debate_r2_mixed_random_mid",32 "validation_G_debate_r4_mixed_random_mid",33 "validation_G_fullview_r2_mixed_random_mid",34 "validation_H_debate_skeptic_r2_mixed_random_mid",35 "validation_I_placebo_r2_mixed_random_mid",36 "validation_J_cost_matched_mixed_random_mid"37 ],38 "truth": "data/processed/prospective_bundle/analytic.parquet",39 "split": "prospective",40 "n_prediction_rows": 7932,41 "n_truth": 1182,42 "B": 2000,43 "seed": 20261027,44 "main_committee": "mixed_random",45 "main_aggregation": "median",46 "min_cell": 30,47 "arms": [48 "A_hedonic",49 "B_hierarchical",50 "C_catboost",51 "C_gbm",52 "C_xgboost",53 "D_avm",54 "E_single_llm",55 "F_independent",56 "G_debate_r2"57 ],58 "fake_included": false,59 "note": ""60}