{"id":"W2948937890","doi":"10.48550/arxiv.1903.03176","title":"MinAtar: An Atari-Inspired Testbed for Thorough and Reproducible Reinforcement Learning Experiments","year":2019,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":42,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Reinforcement learning; Testbed; Computer science; Representation (politics); Breakout; Artificial intelligence; Set (abstract data type); Human–computer interaction; Toolbox; 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":[0.0007138331,0.0005057962,0.0005111899,0.000292453,0.0003589001,0.0003602437,0.001956046,0.0003394531,0.00001888454],"category_scores_gemma":[0.0001909326,0.0006028981,0.0001704323,0.0003387648,0.0001110009,0.001028785,0.003086825,0.0006748926,0.00004747968],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002858322,"about_ca_system_score_gemma":0.0002091302,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001646654,"about_ca_topic_score_gemma":0.000004185508,"domain_scores_codex":[0.9963516,0.0001620883,0.0004041634,0.002277078,0.000206264,0.0005987948],"domain_scores_gemma":[0.996111,0.0001609176,0.0005671998,0.002675385,0.0002544219,0.0002311055],"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":[0.00005672743,0.00004060113,0.003478467,0.000164166,0.00008791967,0.00002996369,0.0007686754,0.9747337,0.0001519425,0.01993597,0.0001644789,0.0003873843],"study_design_scores_gemma":[0.001059176,0.0005373122,0.0004282623,0.0001514263,0.00007408286,0.000003874954,0.0002200519,0.99133,0.0009345397,0.001047793,0.003513999,0.000699522],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08227202,0.00008714031,0.9128402,0.00006143952,0.0007601437,0.001182913,0.000003112281,0.0003779405,0.002415065],"genre_scores_gemma":[0.9681041,0.0001172885,0.0162145,0.0001009813,0.00009610628,0.000008395828,0.0001344422,0.00004602549,0.01517818],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8966258,"threshold_uncertainty_score":0.9996423,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1109329489629851,"score_gpt":0.2332999492194836,"score_spread":0.1223670002564985,"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."}}