{"id":"W4407830460","doi":"10.1109/jbhi.2025.3544560","title":"GPT-Based Automated Induction: Vulnerability Detection in Medical Software","year":2025,"lang":"en","type":"article","venue":"IEEE Journal of Biomedical and Health Informatics","topic":"Software Reliability and Analysis Research","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Brandon University","funders":"Ningbo Municipal Bureau of Science and Technology","keywords":"Computer science; Software; Vulnerability (computing); Computer security; Operating system","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":[],"consensus_categories":[],"category_scores_codex":[0.005782734,0.00009840952,0.0003749211,0.0006971232,0.000157077,0.0000757501,0.0004495983,0.0001996578,0.00001455849],"category_scores_gemma":[0.001126746,0.00007198591,0.00007769813,0.001419132,0.0002091237,0.0003842678,0.00007421544,0.000769878,0.000002901744],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002257727,"about_ca_system_score_gemma":0.001926305,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006391544,"about_ca_topic_score_gemma":0.00003528881,"domain_scores_codex":[0.996826,0.0001929106,0.001574232,0.00009654015,0.001023961,0.0002863388],"domain_scores_gemma":[0.9983685,0.0003923361,0.0003615548,0.0001979325,0.000246606,0.0004330967],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00002832884,0.0003030654,0.003208549,0.0008685417,0.00002811456,0.00001201523,0.0005841966,0.0001702161,0.000007663214,0.000121288,0.001730916,0.9929371],"study_design_scores_gemma":[0.002082282,0.0007238817,0.02141947,0.0007764799,0.000008648795,0.0001380625,0.0001840136,0.9612634,0.0003409598,0.002548025,0.01036395,0.0001508254],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09960699,0.0002182417,0.8784659,0.02086021,0.0006375188,0.0001042531,0.0000013213,0.00009036367,0.00001520881],"genre_scores_gemma":[0.9572529,0.0004075542,0.0379894,0.004190509,0.0001400525,0.000005452162,0.00000200758,0.000003364966,0.000008826141],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9927863,"threshold_uncertainty_score":0.3417183,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02909851817937087,"score_gpt":0.3674956556795596,"score_spread":0.3383971375001887,"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."}}