{"id":"W4388854508","doi":"10.1109/access.2023.3335841","title":"On the Effectiveness of Activation Noise in Both Training and Inference for Generative Classifiers","year":2023,"lang":"en","type":"article","venue":"IEEE Access","topic":"Neural Networks and Applications","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Huawei Technologies (Canada); Queen's University","funders":"Semnan University; Huawei Technologies; Queen's University; Amirkabir University of Technology; Yonsei University; Korea Advanced Institute of Science and Technology; Harvard University","keywords":"Inference; Computer science; Noise (video); Artificial intelligence; Autoencoder; Regularization (linguistics); Machine learning; Pattern recognition (psychology); Artificial neural network","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0149038,0.00193582,0.001365722,0.0009192598,0.001250183,0.002245547,0.002051766,0.003190338,0.002029373],"category_scores_gemma":[0.06811221,0.001028101,0.001091848,0.0006233741,0.003834494,0.004820032,0.003807497,0.0042994,0.0006270246],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00173314,"about_ca_system_score_gemma":0.002354207,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005304478,"about_ca_topic_score_gemma":0.005398422,"domain_scores_codex":[0.9934029,0.003646882,0.0003293061,0.001178645,0.001032962,0.0004092297],"domain_scores_gemma":[0.9539824,0.03991047,0.001160368,0.003016891,0.001443737,0.0004862303],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000769991,0.000222682,0.006234762,0.0001864036,0.0002050109,0.0001424199,0.000363696,0.8022367,0.007169016,0.05740079,0.001060431,0.1240081],"study_design_scores_gemma":[0.00001838846,0.0001329624,0.0007236643,0.00004900255,0.00003051707,0.000062133,0.00002765351,0.9784464,0.004176993,0.01593708,0.0003782601,0.00001693585],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07813894,0.001514886,0.9127091,0.001705247,0.00007729353,0.0000882528,0.00007129944,0.0006537643,0.005041201],"genre_scores_gemma":[0.8143216,0.0009979641,0.1804567,0.0008426427,0.0001571215,0.0001999327,0.0002589081,0.0003204759,0.002444492],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0149038,"threshold_uncertainty_score":0.07881975,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09950281199325181,"score_gpt":0.3530368816332861,"score_spread":0.2535340696400342,"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."}}