{"id":"W2945563437","doi":"10.48550/arxiv.1905.10259","title":"Dichotomize and Generalize: PAC-Bayesian Binary Activated Deep Neural Networks","year":2019,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université Laval","funders":"","keywords":"Binary number; Computer science; Differentiable function; Artificial neural network; Generalization; Artificial intelligence; Deep neural networks; Activation function; Binary Independence Model; Bayesian probability; Bayesian network; Machine learning; Mathematics","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":[0.0001664421,0.0004222895,0.0004297253,0.000376185,0.0001547039,0.0001991286,0.001460737,0.0004223315,0.00002675922],"category_scores_gemma":[0.00003085596,0.0004792244,0.0001488796,0.0006737232,0.00013719,0.0004444676,0.002300289,0.0006265701,0.000007178064],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001589679,"about_ca_system_score_gemma":0.00006492728,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008169265,"about_ca_topic_score_gemma":0.000005415957,"domain_scores_codex":[0.9977602,0.0001846831,0.000230992,0.001291427,0.00009969228,0.0004330028],"domain_scores_gemma":[0.9978817,0.0001346513,0.000316994,0.001300951,0.0001426117,0.0002230921],"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":[0.00003896713,0.0000446704,0.001749886,0.000033759,0.0000660336,0.0001592853,0.000119601,0.9623428,0.0000463545,0.034655,0.0002040149,0.0005396471],"study_design_scores_gemma":[0.0005412255,0.00007271778,0.0006090714,0.00004561057,0.00004197398,0.0000131401,0.00001453155,0.9892626,0.00009078408,0.008793255,0.00002625166,0.0004888625],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06307699,0.0001121557,0.9342808,0.0001030203,0.0006820679,0.0005975286,0.000006157235,0.0006905264,0.0004507845],"genre_scores_gemma":[0.9864179,0.000177905,0.01261474,0.0001899998,0.00005846486,0.000001859486,0.00003021268,0.00003287409,0.0004760163],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9233409,"threshold_uncertainty_score":0.9997659,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03541511775381253,"score_gpt":0.1810671450932165,"score_spread":0.145652027339404,"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."}}