{"id":"W4415160834","doi":"10.48550/arxiv.2504.10397","title":"Can LLMs Assist Expert Elicitation for Probabilistic Causal Modeling?","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; University of Calgary","keywords":"Expert elicitation; Bayesian network; Probabilistic logic; Entropy (arrow of time); Robustness (evolution); Bayesian probability; Causal model; Causal inference; Transparency (behavior)","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.03855423,0.001507594,0.0009415397,0.002891299,0.0008633232,0.003506418,0.002147099,0.001984105,0.009187815],"category_scores_gemma":[0.23434,0.0007300405,0.001312821,0.001482497,0.001711684,0.005325801,0.005355083,0.002655276,0.002047634],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001989791,"about_ca_system_score_gemma":0.003523755,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001972428,"about_ca_topic_score_gemma":0.004451138,"domain_scores_codex":[0.95711,0.03688309,0.001265071,0.002185813,0.002215265,0.0003406563],"domain_scores_gemma":[0.7264661,0.2443613,0.00883857,0.01284934,0.006512005,0.0009727263],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001659199,0.0007591608,0.02049856,0.004841472,0.0007108183,0.0009521829,0.01524112,0.1346493,0.01484304,0.1773927,0.01660755,0.611845],"study_design_scores_gemma":[0.0002363729,0.0002401103,0.002626514,0.001392704,0.0001550331,0.000372387,0.002503795,0.5614617,0.008123567,0.3992055,0.02352756,0.0001546654],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02599334,0.0004140591,0.9633102,0.003589564,0.00007007815,0.0003377784,0.000889755,0.001146286,0.004249039],"genre_scores_gemma":[0.3698973,0.0003834537,0.6253822,0.001005981,0.00008940509,0.0009534133,0.001240277,0.0001579612,0.0008898993],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.03855423,"threshold_uncertainty_score":0.2038966,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1132032458476235,"score_gpt":0.3423743979090132,"score_spread":0.2291711520613897,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}