{"id":"W4393190288","doi":"10.1101/2024.03.25.586710","title":"Have AI-Generated Texts from LLM Infiltrated the Realm of Scientific Writing? A Large-Scale Analysis of Preprint Platforms","year":2024,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University; Impact","funders":"","keywords":"Realm; Preprint; Automatic summarization; Citation; Upload; Computer science; Quality (philosophy); Scale (ratio); Artificial intelligence; World Wide Web; History; Epistemology; Geography; Philosophy; Cartography; Archaeology","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","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.008439759,0.0003246215,0.0004944119,0.009263084,0.001403413,0.004442456,0.0007599113,0.0007483852,0.002146055],"category_scores_gemma":[0.08171009,0.0001946874,0.0004315768,0.009141513,0.001656524,0.003920674,0.003150712,0.001024467,0.001318234],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008715205,"about_ca_system_score_gemma":0.0009254612,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001262169,"about_ca_topic_score_gemma":0.001768021,"domain_scores_codex":[0.9917714,0.00288476,0.001131281,0.001383504,0.002478036,0.0003508981],"domain_scores_gemma":[0.8203323,0.1114743,0.0398502,0.0111585,0.01347888,0.003705886],"domain_codex":null,"domain_gemma":"reporting","domain_candidate":"reporting","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0005945601,0.0002387567,0.8105127,0.001782038,0.0002577207,0.001690163,0.0406905,0.0008999969,0.01369991,0.003782069,0.01078854,0.1150629],"study_design_scores_gemma":[0.00002106727,0.0001343958,0.9393039,0.0003010364,0.0001072807,0.0008441923,0.01456839,0.00454762,0.004786578,0.001620468,0.03367515,0.00009007869],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9883056,0.001094241,0.002024484,0.0007482047,0.0001205603,0.00006328105,0.003336991,0.0002294867,0.004077359],"genre_scores_gemma":[0.9912435,0.0003965994,0.002147679,0.0001349844,0.0003129769,0.0001043564,0.004135945,0.0001236509,0.001400245],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9992516,"threshold_uncertainty_score":0.04463428,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04982422167562414,"score_gpt":0.3290977303638082,"score_spread":0.2792735086881841,"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."}}