{"id":"W4385644074","doi":"10.48550/arxiv.2308.02001","title":"Memory capacity of two layer neural networks with smooth activations","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Machine Learning and ELM","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Artificial neural network; Algebraic number; Rank (graph theory); Computer science; Jacobian matrix and determinant; Hadamard transform; Dimension (graph theory); Layer (electronics); Mathematics; Upper and lower bounds; Discrete mathematics; Theoretical computer science; Algorithm; Applied mathematics; Pure mathematics; Artificial intelligence; Combinatorics; Mathematical analysis","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":[],"consensus_categories":[],"category_scores_codex":[0.0002173358,0.0002364659,0.0002910103,0.0002275884,0.0001515188,0.00006027328,0.001243088,0.000145873,0.00001439485],"category_scores_gemma":[0.00002522187,0.0002352653,0.0001304558,0.0007485913,0.00009603277,0.0002305462,0.001089716,0.0006896147,0.00001587822],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006794035,"about_ca_system_score_gemma":0.00009555159,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001471493,"about_ca_topic_score_gemma":0.0002342574,"domain_scores_codex":[0.9985568,0.0001665642,0.0001483264,0.0007317336,0.0001043163,0.0002922499],"domain_scores_gemma":[0.9982744,0.000153011,0.0002746125,0.001050275,0.0001342061,0.0001134253],"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":[0.00001413233,0.000042402,0.005800872,0.00003014867,0.0000604102,0.00005775194,0.0001996126,0.9757786,0.00000983667,0.01717159,0.0001980922,0.0006365899],"study_design_scores_gemma":[0.0003882337,0.00004939787,0.01029681,0.00005973659,0.00004096356,0.000002550404,0.00002814178,0.9846272,0.00004908184,0.004113335,0.00007519677,0.000269361],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4442177,0.000004717202,0.5538229,0.0001389226,0.0002917976,0.0001069361,0.000004212832,0.0002661415,0.001146766],"genre_scores_gemma":[0.9959556,0.00001109048,0.001808245,0.00005719533,0.0000776196,7.93582e-7,0.00001014147,0.00001990275,0.002059348],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5520146,"threshold_uncertainty_score":0.9593841,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09508457912140833,"score_gpt":0.2061784418389984,"score_spread":0.11109386271759,"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."}}