{"id":"W4399210824","doi":"10.2196/56780","title":"Potential Roles of Large Language Models in the Production of Systematic Reviews and Meta-Analyses","year":2024,"lang":"en","type":"article","venue":"Journal of Medical Internet Research","topic":"Clinical practice guidelines implementation","field":"Medicine","cited_by":59,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University; Impact","funders":"","keywords":"Preprint; Meta-analysis; Production (economics); Computer science; Linguistics; Economics; Philosophy; World Wide Web; Medicine","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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.0501167,0.00007102689,0.0008421011,0.0003717373,0.000009236964,0.00002713543,0.0002558517,0.00007458674,0.0004517356],"category_scores_gemma":[0.04415826,0.00003124648,0.0002781663,0.0003572338,0.0001238637,0.0001812316,0.00008861652,0.0009455482,0.000003559573],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003402007,"about_ca_system_score_gemma":0.0001978777,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001931055,"about_ca_topic_score_gemma":0.0001505942,"domain_scores_codex":[0.9926515,0.002203367,0.002252682,0.0001246068,0.002629403,0.0001384609],"domain_scores_gemma":[0.9957857,0.002973574,0.0004398058,0.0002133859,0.0004921026,0.00009546184],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"systematic_review","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.004510439,0.007641562,0.0009862696,0.4088937,0.03341981,0.006990643,0.08076848,0.0002416206,0.06781562,0.01197332,0.3155878,0.06117079],"study_design_scores_gemma":[0.01216595,0.01415329,0.002718967,0.1677279,0.1148511,0.02196889,0.2233213,0.3846629,0.02899404,0.01338851,0.01519311,0.0008539109],"study_design_candidate":"systematic_review","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.822836,0.1054969,0.005670947,0.06382009,0.000322527,0.001688448,0.000006369632,0.000005044468,0.0001536552],"genre_scores_gemma":[0.9958566,0.003240598,0.0002743411,0.0001534075,0.0002358501,0.00001376256,0.000002002327,0.00000663488,0.0002167332],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3844213,"threshold_uncertainty_score":0.9781048,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6880497700622092,"score_gpt":0.6641778485082145,"score_spread":0.02387192155399476,"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."}}