{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00434186,0.001641105,0.001403269,0.0004277859,0.0005835991,0.0009514631,0.001770946,0.001393691,0.001867401],"category_scores_gemma":[0.02060096,0.000682104,0.0006813243,0.0003393471,0.002107244,0.002131707,0.002090831,0.003597576,0.0004389347],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001490895,"about_ca_system_score_gemma":0.001348814,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004941412,"about_ca_topic_score_gemma":0.003207004,"domain_scores_codex":[0.9981255,0.0005643452,0.0001414812,0.0004847386,0.0004341041,0.0002498601],"domain_scores_gemma":[0.9922239,0.004801912,0.0007284813,0.001388527,0.0006254258,0.000231756],"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":[0.0002272184,0.0002094032,0.002271263,0.0001256073,0.0001016234,0.0000904911,0.0001182989,0.8884044,0.007064402,0.007972267,0.001060188,0.09235483],"study_design_scores_gemma":[0.00002057063,0.00012417,0.0002637627,0.00001148791,0.00001120884,0.00001844542,0.000007242031,0.9918251,0.001690863,0.005774422,0.0002452703,0.000007488397],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1352473,0.0008179937,0.8562963,0.0006850411,0.0001163519,0.0001187295,0.0000587905,0.002351208,0.004308279],"genre_scores_gemma":[0.9335115,0.0001800123,0.06456438,0.0002833161,0.00004952558,0.0001301946,0.00006364513,0.0001388442,0.001078593],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004941412,"threshold_uncertainty_score":0.02296221,"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."}}