{"id":"W4288391571","doi":"10.1109/tse.2022.3194640","title":"FalsifAI: Falsification of AI-Enabled Hybrid Control Systems Guided by Time-Aware Coverage Criteria","year":2022,"lang":"en","type":"article","venue":"IEEE Transactions on Software Engineering","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":24,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Correctness; Robustness (evolution); Artificial neural network; Notation; Context (archaeology); Hybrid system; Semantics (computer science); Artificial intelligence; Model checking; Theoretical computer science; Programming language; Machine learning","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[{"model":"gemma","categories":[],"domain":null,"study_design":"not_applicable","genre":"methods","about_ca_system":false,"about_ca_topic":false,"confidence":"low","status":"direct model label, unvalidated"},{"model":"gpt","categories":[],"domain":null,"study_design":"design_other","genre":"software","about_ca_system":false,"about_ca_topic":false,"confidence":"high","status":"direct model label, unvalidated"}],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0004648542,0.0002604632,0.00037107,0.000281272,0.0003528175,0.0001061889,0.0007641048,0.00005667598,0.0001235544],"category_scores_gemma":[0.0000521174,0.0003127471,0.0001369198,0.0004923893,0.00001894972,0.0004987415,0.00001301929,0.0005922528,0.00001876609],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002541339,"about_ca_system_score_gemma":0.0000723566,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00007117319,"about_ca_topic_score_gemma":2.370359e-7,"domain_scores_codex":[0.9979693,0.0001856545,0.0005029825,0.0004644881,0.0005321717,0.0003453498],"domain_scores_gemma":[0.9986311,0.000340911,0.0001625612,0.0006483893,0.0001189261,0.00009811047],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001831195,0.0001034589,0.00001628393,0.0000588056,0.00005870197,0.00001248129,0.0001312084,0.9905532,0.00651051,0.000116346,0.00117151,0.001249156],"study_design_scores_gemma":[0.001033117,0.0001277793,0.00002743778,0.00003809905,0.00002911339,0.00004329429,0.00001746136,0.9912822,0.005065442,0.00001405076,0.002006992,0.0003149603],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.007274268,0.00007239712,0.9895239,0.00009025399,0.001648385,0.000398678,0.0002070067,0.0007689262,0.00001611867],"genre_scores_gemma":[0.9913856,0.000004150986,0.007828046,0.00008789248,0.00004044886,0.0002063693,0.0000190282,0.0000513203,0.0003771259],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9841114,"threshold_uncertainty_score":0.9999325,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.007528728385960206,"score_gpt":0.225826501456591,"score_spread":0.2182977730706308,"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."}}