{"id":"W4393148616","doi":"10.1609/aaai.v38i5.28228","title":"Adversarial Attacks on the Interpretation of Neuron Activation Maximization","year":2024,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University; Mila - Quebec Artificial Intelligence Institute","funders":"Fonds de recherche du Québec – Nature et technologies","keywords":"Interpretation (philosophy); Maximization; Adversarial system; Computer science; Artificial intelligence; Computer security; Mathematics; Mathematical optimization; Programming language","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0007179992,0.0001604277,0.0001543027,0.0001716507,0.0001093672,0.0002695336,0.001205,0.00005789055,0.00003686856],"category_scores_gemma":[0.0008490662,0.0001040999,0.0001181269,0.000883965,0.0001586501,0.0006085943,0.0002260916,0.0002657537,0.00003324201],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005988937,"about_ca_system_score_gemma":0.000131201,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001616814,"about_ca_topic_score_gemma":0.000001370007,"domain_scores_codex":[0.9983458,0.000067886,0.0004636044,0.0003842659,0.0005836768,0.0001547481],"domain_scores_gemma":[0.9983031,0.0006968534,0.0002868679,0.0002342252,0.0004500153,0.00002898931],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00007280768,0.00005236299,0.00001004107,0.00003801479,0.00001310884,1.083007e-7,0.001442791,0.01646725,0.01657319,0.876586,0.0001124236,0.08863197],"study_design_scores_gemma":[0.000008879416,0.0001033505,0.0001288574,0.0002411319,0.000006313739,7.357967e-7,0.00009326284,0.5649578,0.2471588,0.1872112,0.00002553442,0.00006414072],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1803348,0.00001070089,0.7994686,0.01053809,0.00128819,0.0005620774,0.000005736396,0.0001050432,0.007686718],"genre_scores_gemma":[0.9974639,0.00000937651,0.002159112,0.0002020687,0.00007310905,0.00001883399,9.9421e-7,0.00001041406,0.00006212458],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8171292,"threshold_uncertainty_score":0.424507,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07681338516569006,"score_gpt":0.3261770996907368,"score_spread":0.2493637145250467,"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."}}