{"id":"W7039154034","doi":"","title":"Limitations of information : theoretic generalization bounds for gradient descent methods in stochastic convex optimization","year":2023,"lang":"en","type":"article","venue":"KTH Publication Database DiVA (KTH Royal Institute of Technology)","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Vetenskapsrådet; University of Toronto","keywords":"Minimax; Generalization; Stochastic gradient descent; Gradient descent; Mutual information; Regular polygon; Convex optimization; Convex function","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001155042,0.0002333663,0.0003514286,0.003299701,0.0001559011,0.00009511631,0.001153385,0.000220406,0.000009469899],"category_scores_gemma":[0.004840326,0.000249699,0.00008015257,0.004891289,0.0003585583,0.001950879,0.0004146631,0.0001534325,0.000005510652],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001515281,"about_ca_system_score_gemma":0.0002426953,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002831027,"about_ca_topic_score_gemma":0.000007836607,"domain_scores_codex":[0.9976309,0.00007481257,0.001141164,0.0004386757,0.0003544534,0.0003600084],"domain_scores_gemma":[0.9964871,0.0002149543,0.0008371681,0.001062654,0.001316645,0.00008149043],"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":[0.00001116772,0.0001036557,0.00009538179,0.00007088407,0.0000175684,1.8993e-7,0.0002047508,0.3286742,0.0001741801,0.6550891,0.0009427716,0.01461612],"study_design_scores_gemma":[0.0007012177,0.0001242594,0.0002540781,0.00009836949,0.00002709456,0.000002640039,0.00007922822,0.9831318,0.003962785,0.008880677,0.002512901,0.0002249648],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0008226661,0.00004347707,0.9932926,0.002635252,0.000448322,0.001481131,0.0001765418,0.0008310298,0.0002689641],"genre_scores_gemma":[0.09498134,0.00009160149,0.901021,0.00009870068,0.00001301747,0.0008552489,0.002890169,0.00001780401,0.00003114216],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.6544576,"threshold_uncertainty_score":0.9999955,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06181125478238244,"score_gpt":0.3177407573990477,"score_spread":0.2559295026166653,"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."}}