{"id":"W4382203111","doi":"10.1609/aaai.v37i8.26186","title":"Fast Convergence in Learning Two-Layer Neural Networks with Separable Data","year":2023,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"National Science Foundation","keywords":"Overfitting; Generalization; Convergence (economics); Separable space; Applied mathematics; Stability (learning theory); Mathematics; Artificial neural network; Gradient descent; Stochastic gradient descent; Exponential function; Exponential stability; Computer science; Mathematical optimization; Algorithm; Artificial intelligence; Mathematical analysis; Nonlinear system; Machine learning","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":[0.0007610173,0.0002311905,0.0002586636,0.0002565682,0.0001781479,0.0002413627,0.003517378,0.00006904681,0.00003384833],"category_scores_gemma":[0.0003723772,0.0001750135,0.00003924307,0.002448842,0.0002347149,0.0007660076,0.001090065,0.0004573332,0.00004256096],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003640892,"about_ca_system_score_gemma":0.00006983738,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006824546,"about_ca_topic_score_gemma":0.0000240238,"domain_scores_codex":[0.9977691,0.00002779007,0.0004911568,0.0007232861,0.0005071913,0.0004815055],"domain_scores_gemma":[0.9984776,0.000144527,0.0003311214,0.0005716543,0.0003943267,0.00008076166],"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":[0.00009707829,0.0001167722,0.006169556,0.00003453423,0.00001797912,0.000004862312,0.001545812,0.29201,0.004324442,0.667458,0.0003618806,0.02785906],"study_design_scores_gemma":[0.00003637424,0.0001449135,0.0002368368,0.0001623726,0.000005416391,0.000005527491,0.0003309909,0.9607296,0.02580609,0.01232563,0.00001467705,0.0002015993],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08028902,0.00002007937,0.9104862,0.002268122,0.0005503462,0.0008303134,0.000005200458,0.000716585,0.004834105],"genre_scores_gemma":[0.9926726,0.00003410392,0.006888046,0.00008712513,0.00003388403,0.0000391492,0.000003327122,0.00001723844,0.000224536],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9123836,"threshold_uncertainty_score":0.7136842,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1136395192751324,"score_gpt":0.3207995889363792,"score_spread":0.2071600696612468,"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."}}