{"id":"W4391420895","doi":"10.1109/host55342.2024.10545398","title":"Data-Oblivious ML Accelerators Using Hardware Security Extensions","year":2024,"lang":"en","type":"preprint","venue":"","topic":"Security and Verification in Computing","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Encryption; Confidentiality; Overhead (engineering); Field-programmable gate array; Embedded system; Cryptography; Side channel attack; Homomorphic encryption; Operating system; Computer security","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":["metaepi_narrow","scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.0008089821,0.0004151608,0.0004175238,0.0003136235,0.0003534915,0.001868731,0.005368266,0.0003929642,0.00009933858],"category_scores_gemma":[0.0001378613,0.0004042007,0.0001597061,0.0006459647,0.00006849239,0.0004206278,0.02695813,0.001580954,0.0002529051],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001385468,"about_ca_system_score_gemma":0.0007471748,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002963792,"about_ca_topic_score_gemma":0.00004226081,"domain_scores_codex":[0.9962328,0.0001512793,0.0006122814,0.00200381,0.0005507192,0.0004491048],"domain_scores_gemma":[0.9944041,0.0001764174,0.0002215946,0.004726289,0.0002725298,0.0001990412],"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":[0.000009752202,0.0006775091,0.0003482414,0.00224983,0.0007111912,0.0005989016,0.01174321,0.01142864,0.0005000089,0.6832678,0.242719,0.04574582],"study_design_scores_gemma":[0.00006013514,0.000007448843,0.0000485861,0.0002835084,0.00004183496,0.00005317021,0.00006973194,0.91743,0.0002158751,0.06327576,0.01800728,0.0005066073],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04478632,0.003234309,0.9239417,0.002904488,0.01381882,0.000751464,0.0004101838,0.003121707,0.007030948],"genre_scores_gemma":[0.8498461,0.0001044301,0.1476264,0.0009774507,0.0009636871,0.00001656643,0.0002198781,0.00005244842,0.0001929903],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9060014,"threshold_uncertainty_score":0.999841,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1720966076992857,"score_gpt":0.3656778762821333,"score_spread":0.1935812685828476,"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."}}