{"id":"W7103418229","doi":"","title":"Can Contemporary Large Language Models Provide the Domain Knowledge Needed for Causal Inference? Evaluating Automated Causal Graph Discovery Through an ASCVD Case Study","year":2025,"lang":"en","type":"article","venue":"DOAJ (DOAJ: Directory of Open Access Journals)","topic":"Advanced Causal Inference Techniques","field":"Mathematics","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute of Population and Public Health","funders":"","keywords":"Causal inference; Directed acyclic graph; Inference; Consistency (knowledge bases); Population; Public health; Baseline (sea); Population health; Epidemiology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.06304328,0.001087104,0.0009673808,0.002216195,0.001086612,0.002686874,0.003040118,0.002466762,0.00407087],"category_scores_gemma":[0.3439479,0.00111519,0.002094426,0.002083273,0.00229893,0.005862817,0.002631759,0.00341976,0.000454429],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002974037,"about_ca_system_score_gemma":0.003647537,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007836394,"about_ca_topic_score_gemma":0.01464271,"domain_scores_codex":[0.9528906,0.04071421,0.001556731,0.002186859,0.002167803,0.0004837497],"domain_scores_gemma":[0.3230137,0.6488119,0.006254802,0.01610891,0.005041034,0.0007697189],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.005347824,0.002464946,0.1502063,0.002976906,0.001355901,0.0024021,0.007049222,0.5003262,0.002279909,0.06097116,0.01247256,0.2521471],"study_design_scores_gemma":[0.000918624,0.0006111616,0.005649087,0.0003663444,0.0003274385,0.0003037771,0.001388008,0.8798231,0.00223259,0.1003039,0.008006564,0.00006920895],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5944991,0.001597118,0.384634,0.00862811,0.0002300203,0.001338259,0.003760303,0.001182045,0.004131084],"genre_scores_gemma":[0.7681192,0.0004141532,0.2270071,0.0006797695,0.00008325672,0.0009156297,0.002273756,0.0001157062,0.0003914284],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9369567,"threshold_uncertainty_score":0.3334087,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5429744453255869,"score_gpt":0.6576209819113161,"score_spread":0.1146465365857292,"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."}}