{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003255244,0.0002358503,0.0005063012,0.0001263172,0.0002943745,0.0001675838,0.001406156,0.0001316937,5.383264e-7],"category_scores_gemma":[0.001705696,0.0001527138,0.0002944162,0.0006636817,0.0005666388,0.001200856,0.0009008098,0.0006446237,2.810507e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001116709,"about_ca_system_score_gemma":0.0001070341,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002725737,"about_ca_topic_score_gemma":0.00001170516,"domain_scores_codex":[0.9978349,0.00007149514,0.0008835635,0.0005145207,0.000445117,0.0002503819],"domain_scores_gemma":[0.9976686,0.000885298,0.0006038236,0.0003705549,0.0004158249,0.00005596348],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00008119862,0.00005633096,0.4795447,0.01262401,0.000228236,2.807837e-7,0.02225608,0.002825481,0.0001550691,0.4634602,0.00002640925,0.018742],"study_design_scores_gemma":[0.0004285643,0.0002576074,0.005055828,0.003451323,0.0002599115,0.0000288624,0.0008893371,0.9357285,0.002632151,0.04956777,0.001381156,0.0003189752],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4902118,0.000770896,0.505082,0.001320604,0.0006401636,0.001403779,0.00003067831,0.0002107507,0.0003293272],"genre_scores_gemma":[0.9833069,0.00000164509,0.01653076,0.000008621975,0.00005546286,0.00005909804,0.000001973338,0.0000180347,0.00001754298],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9329031,"threshold_uncertainty_score":0.6227489,"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."}}