{"id":"W3010862467","doi":"10.48550/arxiv.2003.07417","title":"Improving Performance in Reinforcement Learning by Breaking Generalization in Neural Networks","year":2020,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Reinforcement learning; Computer science; Artificial intelligence; Generalization; Artificial neural network; Scalability; 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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0003420476,0.0003431157,0.0003405932,0.0003451003,0.0001273502,0.0002062577,0.001387743,0.0002689308,0.000009688617],"category_scores_gemma":[0.00005465152,0.0004483062,0.00007585772,0.001013582,0.00004183584,0.0007189517,0.002214864,0.001430237,0.000009845211],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005362781,"about_ca_system_score_gemma":0.00009514564,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003787921,"about_ca_topic_score_gemma":0.0000295051,"domain_scores_codex":[0.9977328,0.0001837242,0.0004306124,0.0009685408,0.0001455856,0.0005387734],"domain_scores_gemma":[0.9987372,0.00005813843,0.0004684952,0.0005624807,0.00006287729,0.0001108628],"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":[0.00001496249,0.000009570971,0.03660345,0.00007998082,0.00001059961,0.00007515311,0.0002894138,0.9595113,0.0000313467,0.002295549,0.00001881036,0.001059863],"study_design_scores_gemma":[0.0005155833,0.00008407389,0.00178308,0.0001320796,0.00001068903,0.0000014858,0.00002914795,0.9968733,0.0000245146,0.00004028596,0.00007972252,0.0004260922],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09268743,0.00004666092,0.9057804,0.00006070953,0.0003555008,0.0003050748,2.29762e-7,0.00017538,0.0005886355],"genre_scores_gemma":[0.9982737,0.0001873405,0.000773538,0.000147881,0.00005695809,0.000001910633,0.00005476268,0.00002549861,0.0004783617],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9055863,"threshold_uncertainty_score":0.9997969,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03919345243744466,"score_gpt":0.1797400113986028,"score_spread":0.1405465589611581,"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."}}