{"id":"W4383873581","doi":"10.20944/preprints202307.0192.v2","title":"Examining the Potential of Generative Language Models for Aviation Safety Analysis: Case Study and Insights using the Aviation Safety Reporting System (ASRS)","year":2023,"lang":"en","type":"preprint","venue":"Preprints.org","topic":"Occupational Health and Safety Research","field":"Health Professions","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"General Fusion (Canada)","funders":"","keywords":"Aviation; Computer science; Crew; Similarity (geometry); Aviation safety; Aviation accident; Process (computing); Generative grammar; Risk analysis (engineering); Operations research; Aeronautics; Engineering; Artificial intelligence; Business","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":["sts"],"consensus_categories":[],"category_scores_codex":[0.01303589,0.0003373074,0.0008904618,0.000390323,0.00287732,0.0000250641,0.0003816596,0.0004077627,0.00003177783],"category_scores_gemma":[0.002022891,0.0002261704,0.0002388968,0.0006994112,0.00008859031,0.0001678924,0.001530341,0.001468111,0.0000127184],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006525049,"about_ca_system_score_gemma":0.001018242,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009637287,"about_ca_topic_score_gemma":0.002487667,"domain_scores_codex":[0.9909066,0.003363704,0.00333473,0.001002399,0.0008531158,0.0005394712],"domain_scores_gemma":[0.9906995,0.002490981,0.004181923,0.001338233,0.001140984,0.000148361],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0008265378,0.0001461579,0.68731,0.002533891,0.001667451,0.000159885,0.1191717,0.1854474,0.0006058856,0.001205359,0.000004932281,0.0009208288],"study_design_scores_gemma":[0.000668137,0.00003846389,0.4373055,0.0003387724,0.000840038,0.00001957648,0.1280862,0.4319495,0.00004841008,0.0005111037,0.000003916365,0.0001903829],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9328017,0.0001668178,0.05763217,0.0002710698,0.0006595274,0.00788695,0.0002290578,0.0001066963,0.0002460076],"genre_scores_gemma":[0.997081,0.00006281731,0.0004439066,0.00004332159,0.0004455195,0.001260375,0.0003381174,0.00005364362,0.0002712881],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2500045,"threshold_uncertainty_score":0.9984208,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4517070882426561,"score_gpt":0.5252554654343733,"score_spread":0.0735483771917172,"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."}}