{"id":"W4407394336","doi":"10.15514/ispras-2024-36(5)-9","title":"Is AI Interpretability Safe: the Relationship between Interpretability and Security of Machine Learning Models","year":2024,"lang":"en","type":"article","venue":"Proceedings of the Institute for System Programming of RAS","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Air Canada","funders":"","keywords":"Interpretability; Artificial intelligence; Computer science; 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":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.03746983,0.001346143,0.002030728,0.003167873,0.001979529,0.00962448,0.002546909,0.0050156,0.003752436],"category_scores_gemma":[0.2856463,0.001205005,0.002461066,0.001663899,0.01611867,0.0213107,0.007090984,0.01069429,0.0008034379],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002449014,"about_ca_system_score_gemma":0.002224459,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001113801,"about_ca_topic_score_gemma":0.0007925886,"domain_scores_codex":[0.9454122,0.03274743,0.003495684,0.007691516,0.008740497,0.001912697],"domain_scores_gemma":[0.4702732,0.3870649,0.03559005,0.09234164,0.01193596,0.002794328],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.001494165,0.0004111711,0.05506507,0.000959723,0.001309181,0.0008870444,0.005190623,0.2080302,0.0058865,0.6008292,0.005996499,0.1139405],"study_design_scores_gemma":[0.00005379946,0.0001534778,0.002167303,0.0001844379,0.00008576478,0.0003853797,0.0004140242,0.1885463,0.002726305,0.8031878,0.002035445,0.00005993122],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1664081,0.001804186,0.7990683,0.0190279,0.0002357083,0.0002180421,0.0007453702,0.0009308704,0.01156146],"genre_scores_gemma":[0.9548787,0.0004902216,0.04176333,0.001000774,0.000323967,0.0001111287,0.0004635004,0.0002741233,0.0006941687],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.03746983,"threshold_uncertainty_score":0.1981618,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03439145558746889,"score_gpt":0.2926161792302561,"score_spread":0.2582247236427872,"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."}}