{"id":"W6910226933","doi":"10.48448/x6dd-pp74","title":"The Effect of Multi-step Methods on Overestimation in Deep Reinforcement Learning","year":2020,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Reinforcement learning; Deep learning; Representation (politics); Artificial neural network; Temporal difference learning; Deep neural networks","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":[0.005328757,0.0003697221,0.0004746731,0.000817636,0.0002329445,0.0001058064,0.0009856569,0.0001592955,0.0002051518],"category_scores_gemma":[0.004286809,0.0002457896,0.00007703657,0.001758135,0.0009479737,0.0001235724,0.0002822784,0.000663657,0.0006148597],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004305186,"about_ca_system_score_gemma":0.0002169458,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005183754,"about_ca_topic_score_gemma":0.0006658142,"domain_scores_codex":[0.9963868,0.000789191,0.0005273562,0.0006350158,0.001148886,0.0005127229],"domain_scores_gemma":[0.9974563,0.001028369,0.000775485,0.0005621348,0.00005587791,0.0001218318],"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":[0.0003538734,0.00009955173,0.001246518,0.0004145336,0.0001012707,0.00001561923,0.0008250893,0.5837963,0.02387612,0.005887242,0.01076195,0.3726219],"study_design_scores_gemma":[0.0009613553,0.0007605777,0.0001710872,0.0003590571,0.00002939539,0.000001320291,0.00007563461,0.9532747,0.004308601,0.00001761556,0.03978024,0.0002603628],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0003636953,0.0009093964,0.5074776,0.0002759916,0.001272005,0.004224008,0.00001598156,0.0007657385,0.4846956],"genre_scores_gemma":[0.2273999,0.0002081743,0.582857,0.000318468,0.000644244,0.0003652157,0.0002716872,0.002619171,0.1853162],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.3723615,"threshold_uncertainty_score":0.9999995,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0332757155414774,"score_gpt":0.3834781824723748,"score_spread":0.3502024669308974,"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."}}