{"id":"W3165994454","doi":"","title":"Reinforcement Learning as One Big Sequence Modeling Problem","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":17,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Reinforcement learning; Computer science; Sequence (biology); Transformer; Artificial intelligence; Markov chain; Sequence learning; Markov decision process; Machine learning; Markov process; Mathematics; Engineering","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":[0.000521717,0.0005165364,0.0005196509,0.0003710915,0.0003909589,0.0006120431,0.002651279,0.000407518,0.00007861171],"category_scores_gemma":[0.000122505,0.0006750432,0.0002849601,0.0008780895,0.00009166819,0.0006661874,0.005272694,0.001795248,0.0001665033],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006059037,"about_ca_system_score_gemma":0.0006803761,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005614469,"about_ca_topic_score_gemma":0.00001306846,"domain_scores_codex":[0.9965367,0.0002392149,0.0004881837,0.001628279,0.0003714005,0.0007361824],"domain_scores_gemma":[0.9969291,0.00009078137,0.0005136204,0.001727912,0.0004620945,0.0002764609],"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":[0.000009556486,0.00002771924,0.0002678148,0.0001494269,0.0001325367,0.0003221605,0.0005886379,0.9533265,0.0001151015,0.04452692,0.00001043306,0.0005232586],"study_design_scores_gemma":[0.0003599229,0.0001123742,0.000008865636,0.0004772363,0.00007368211,0.000007187196,0.0001979058,0.9935911,0.0001621847,0.004130645,0.0001754691,0.000703424],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03559203,0.00005431515,0.9510863,0.0001087561,0.0005572976,0.0004153374,6.046955e-7,0.0005687414,0.01161656],"genre_scores_gemma":[0.9760955,0.0004882279,0.01435269,0.000187402,0.0000992047,0.000002855667,0.00005322665,0.00003981305,0.008681077],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9405035,"threshold_uncertainty_score":0.9995701,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1422045581090734,"score_gpt":0.2102997366343601,"score_spread":0.06809517852528674,"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."}}