{"id":"W2817967199","doi":"10.1613/jair.1.12463","title":"The Bottleneck Simulator: A Model-Based Deep Reinforcement Learning Approach","year":2020,"lang":"en","type":"preprint","venue":"Journal of Artificial Intelligence Research","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Université de Montréal; Polytechnique Montréal; Mila - Quebec Artificial Intelligence Institute","funders":"Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs; Compute Canada; Amazon Web Services; Nuance Foundation; Canadian Institute for Advanced Research; Nvidia","keywords":"Bottleneck; Reinforcement learning; Computer science; Variance (accounting); Task (project management); Obstacle; Artificial intelligence; State space; Engineering; Statistics; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"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","scholarly_communication","open_science","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.009576301,0.0004262276,0.0006478074,0.0006482528,0.00111442,0.002265962,0.006387353,0.0003762155,0.00001831635],"category_scores_gemma":[0.003551013,0.0003169867,0.0005456689,0.001074445,0.0004892864,0.0003852339,0.003134724,0.007386395,0.0001111125],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006292011,"about_ca_system_score_gemma":0.002422702,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001855089,"about_ca_topic_score_gemma":0.00000271287,"domain_scores_codex":[0.990688,0.001084367,0.002111946,0.0006830533,0.004277774,0.001154839],"domain_scores_gemma":[0.992606,0.001715046,0.001338769,0.001286954,0.002542421,0.0005107651],"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":[0.0001223941,0.00005688882,0.00001137095,0.0001118578,0.00009779144,0.00003567083,0.001394734,0.9499118,0.0002208706,0.02496783,0.0002410885,0.02282771],"study_design_scores_gemma":[0.00004903008,0.0005845285,0.000001372625,0.0001730927,0.00001974396,0.000008514786,0.0004864162,0.9741672,0.002746564,0.02030202,0.001170765,0.0002907078],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0002832109,0.0005037952,0.9930928,0.003199608,0.0007393008,0.0005903254,3.460624e-7,0.00006689233,0.001523765],"genre_scores_gemma":[0.9415709,0.0005029519,0.05682543,0.0001066608,0.0006313639,0.00002669632,0.0000047623,0.00005087327,0.0002803302],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9412877,"threshold_uncertainty_score":0.9999282,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2009218619017905,"score_gpt":0.4031124373316161,"score_spread":0.2021905754298257,"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."}}