{"id":"W3211638756","doi":"","title":"An Analysis of Constant Step Size SGD in the Non-convex Regime: Asymptotic Normality and Bias","year":2021,"lang":"en","type":"article","venue":"arXiv (Cornell University)","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Mathematics; Iterated function; Stochastic gradient descent; Constant (computer programming); Applied mathematics; Mathematical optimization; Asymptotic distribution; Convex function; Regular polygon; Rate of convergence; Convex optimization; Markov chain; Computer science; Artificial intelligence; Statistics; Estimator; Mathematical analysis; Artificial neural network","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":[],"consensus_categories":[],"category_scores_codex":[0.0004223526,0.0001030594,0.0002316165,0.0002398797,0.0000650937,0.00005190165,0.0006125867,0.00005304481,0.00002229184],"category_scores_gemma":[0.0001063214,0.00009577333,0.00007329674,0.002898938,0.0001484685,0.0003555447,0.0001527321,0.00009247686,9.846832e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004034443,"about_ca_system_score_gemma":0.0000756023,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001507904,"about_ca_topic_score_gemma":0.0001456101,"domain_scores_codex":[0.9989475,0.0002092816,0.0001714231,0.0004205697,0.00009167587,0.0001595275],"domain_scores_gemma":[0.9984895,0.0003990612,0.0001274405,0.0007656689,0.0001543067,0.00006403073],"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":[0.00002126696,0.0002839939,0.1784521,0.0000218617,0.0002199773,0.0003258733,0.00164472,0.07999638,0.0002085963,0.7384459,0.00003727216,0.0003421214],"study_design_scores_gemma":[0.0003013538,0.000062406,0.06447697,0.00001263149,0.0001526756,0.000005038215,0.0005586792,0.9314675,0.0003883459,0.002438949,0.000009238666,0.0001262521],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3877072,0.00001062023,0.6113925,0.00005934428,0.00002122603,0.00007814568,0.000003423621,0.00003303595,0.0006945181],"genre_scores_gemma":[0.9929196,0.0000274596,0.006852761,0.0001363065,0.000003132379,4.024703e-7,0.0000031643,0.000003054118,0.00005413854],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8514711,"threshold_uncertainty_score":0.3905523,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04739091611543879,"score_gpt":0.1976985429670642,"score_spread":0.1503076268516254,"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."}}