{"id":"W4387144848","doi":"10.2196/48996","title":"Automated Paper Screening for Clinical Reviews Using Large Language Models: Data Analysis Study","year":2023,"lang":"en","type":"article","venue":"Journal of Medical Internet Research","topic":"Meta-analysis and systematic reviews","field":"Decision Sciences","cited_by":177,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; University of Calgary","funders":"","keywords":"Computer science; Workflow; Machine learning; Python (programming language); Artificial intelligence; Natural language processing; Information retrieval; Data science; Data mining; Database; Programming language","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","scholarly_communication","open_science","insufficient_payload"],"consensus_categories":["metaresearch","insufficient_payload"],"category_scores_codex":[0.7778753,0.0002308797,0.005243443,0.002253131,0.0001496465,0.001073965,0.01014097,0.0002261154,0.0153159],"category_scores_gemma":[0.3331864,0.000102344,0.002883209,0.006919451,0.0001326088,0.0006634713,0.002264166,0.001339968,0.000849909],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004562445,"about_ca_system_score_gemma":0.0003033082,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008481838,"about_ca_topic_score_gemma":0.0003093585,"domain_scores_codex":[0.8919721,0.05348621,0.02233562,0.001406955,0.02994941,0.0008496815],"domain_scores_gemma":[0.9493373,0.03123356,0.006808169,0.006983861,0.004469327,0.001167784],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00006242027,0.0004922705,0.02363258,0.00006121532,0.003787317,0.0003210595,0.001440133,0.0002111838,0.000006571081,0.0001092617,0.9240161,0.04585995],"study_design_scores_gemma":[0.0006159284,0.0001950341,0.00250444,0.0001803812,0.0009379392,0.0000201276,0.005266257,0.8848037,3.980844e-7,0.0002577371,0.1051306,0.0000874519],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4726398,0.01014651,0.508602,0.004832344,0.0007861539,0.0023453,0.00009360825,0.00002771603,0.0005265494],"genre_scores_gemma":[0.9673539,0.002052999,0.01745739,0.0006165822,0.001593601,0.00003538889,0.00006900447,0.00004568901,0.01077546],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8845925,"threshold_uncertainty_score":0.999963,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.9651539480355735,"score_gpt":0.7553193395729202,"score_spread":0.2098346084626533,"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."}}