{"id":"W4327919584","doi":"10.1016/j.micpro.2023.104800","title":"EAM: Ensemble of approximate multipliers for robust DNNs","year":2023,"lang":"en","type":"article","venue":"Microprocessors and Microsystems","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"Lakehead University","funders":"Natural Sciences and Engineering Research Council of Canada; Mitacs","keywords":"Computer science; Robustness (evolution); Deep neural networks; Transferability; Exploit; Adversarial system; Multiplier (economics); Range (aeronautics); Deep learning; Artificial intelligence; Computer engineering; Machine learning; Computer security","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":[0.0006330267,0.0002070965,0.0003691549,0.0001999989,0.000259105,0.000125324,0.0006011655,0.0001157583,0.000001422511],"category_scores_gemma":[0.00009748455,0.0001909815,0.0000861754,0.0006074212,0.0000814251,0.0002426506,0.0002971041,0.0001251623,0.00001220443],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002577241,"about_ca_system_score_gemma":0.0000563459,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008664089,"about_ca_topic_score_gemma":0.00001274755,"domain_scores_codex":[0.9984467,0.00003814697,0.0003896665,0.0005211314,0.0001549586,0.0004493874],"domain_scores_gemma":[0.9990247,0.0001495024,0.0002493437,0.0003379465,0.0001548115,0.0000837248],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00007927561,0.00008870541,0.003771689,0.003499545,0.000118811,0.00001757645,0.009641192,0.02282169,0.9183099,0.01184285,0.003751672,0.02605714],"study_design_scores_gemma":[0.003006881,0.000197913,0.0004221703,0.0004118162,0.00004305326,0.00008360465,0.001812633,0.8255002,0.1481127,0.00110427,0.01843168,0.0008730184],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1703281,0.0004398995,0.8276461,0.0001632948,0.0004729458,0.0004793007,0.00002350098,0.0002790653,0.0001677902],"genre_scores_gemma":[0.9324723,0.00005104218,0.06591854,0.00004653356,0.00009806343,0.00007120046,0.00001892004,0.00004091782,0.001282491],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8026786,"threshold_uncertainty_score":0.7788,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02134753520366136,"score_gpt":0.2552095641008111,"score_spread":0.2338620288971498,"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."}}