{"id":"W3194952006","doi":"10.48550/arxiv.2108.06325","title":"Continual Backprop: Stochastic Gradient Descent with Persistent Randomness","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Domain Adaptation and Few-Shot Learning","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Randomness; Initialization; Computer science; Gradient descent; Stochastic gradient descent; Process (computing); Artificial intelligence; Algorithm; Artificial neural network; Mathematics; Statistics","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"],"consensus_categories":[],"category_scores_codex":[0.000296645,0.0003910729,0.0004990538,0.0002206085,0.0002466465,0.0003718017,0.001232075,0.0001791238,0.00006580728],"category_scores_gemma":[0.00004235548,0.0003936838,0.0004301596,0.0005547562,0.0001585871,0.0003209308,0.001182932,0.0006839816,0.00005107577],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002893391,"about_ca_system_score_gemma":0.0003658022,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006724081,"about_ca_topic_score_gemma":0.000056932,"domain_scores_codex":[0.9975027,0.0002641581,0.0002329911,0.001308868,0.000217807,0.000473413],"domain_scores_gemma":[0.9979255,0.0001149684,0.0002978669,0.001000996,0.000357645,0.0003030201],"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":[0.0002477024,0.0002403338,0.0005579773,0.0000980584,0.0004261874,0.001099145,0.002728597,0.9556569,0.000023705,0.03730407,0.00007044621,0.001546916],"study_design_scores_gemma":[0.004946144,0.0002295936,0.00152537,0.0004876573,0.0003063822,0.00006613434,0.003199856,0.9868476,0.00003657028,0.0007134308,0.0005641656,0.00107709],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2104938,0.0001692132,0.7870232,0.0001057985,0.0005543405,0.0003728723,0.000003200279,0.000192492,0.001085035],"genre_scores_gemma":[0.9943059,0.00003130106,0.003416467,0.0001328168,0.00004908242,0.000002899342,0.00002517862,0.00002372604,0.002012625],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7838121,"threshold_uncertainty_score":0.9998515,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06062449877952751,"score_gpt":0.1701160711287008,"score_spread":0.1094915723491733,"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."}}