{"id":"W4416416033","doi":"10.1016/j.jdent.2025.106245","title":"Accuracy of LLMs to retrieve numeric data for meta-analysis in dentistry","year":2025,"lang":"en","type":"article","venue":"Journal of Dentistry","topic":"Meta-analysis and systematic reviews","field":"Decision Sciences","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Limit (mathematics); Outcome (game theory); Risk assessment; MEDLINE","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":["metaresearch"],"category_scores_codex":[0.3036244,0.002764493,0.005034157,0.01679046,0.001134535,0.008014272,0.003144483,0.002079021,0.008048774],"category_scores_gemma":[0.7501447,0.001735111,0.01567336,0.01266063,0.001383051,0.005943545,0.006121835,0.001979267,0.002169681],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00243844,"about_ca_system_score_gemma":0.006070532,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002809681,"about_ca_topic_score_gemma":0.007465146,"domain_scores_codex":[0.6282746,0.2220916,0.105999,0.01779008,0.0248574,0.000987284],"domain_scores_gemma":[0.1436773,0.7819307,0.02727771,0.03355407,0.01301968,0.0005405869],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.007414945,0.0002331537,0.1265252,0.1516058,0.06802656,0.0007245896,0.007972242,0.01213157,0.005181608,0.006248539,0.05284694,0.5610889],"study_design_scores_gemma":[0.00735519,0.003236466,0.236165,0.1056512,0.1473007,0.004132797,0.005096565,0.1520025,0.0339458,0.07372293,0.2289604,0.002430536],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1419264,0.09151623,0.5580725,0.01499462,0.002061422,0.01418481,0.1174563,0.04321275,0.016575],"genre_scores_gemma":[0.4271245,0.008835322,0.5185719,0.002682324,0.0003207644,0.01483455,0.02261255,0.003566213,0.001451936],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6963756,"threshold_uncertainty_score":0.8587551,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.8249960213679443,"score_gpt":0.6056581661013204,"score_spread":0.2193378552666239,"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."}}