{"id":"W3157293568","doi":"10.48550/arxiv.2104.13877","title":"Autoregressive Dynamics Models for Offline Policy Evaluation and\\n Optimization","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Autoregressive model; Computer science; Conditional independence; Covariance; Feed forward; Artificial intelligence; Econometrics; Mathematics; Statistics","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":[0.0003709845,0.0002485079,0.0002598025,0.0003128902,0.0001652764,0.0002640205,0.0007839281,0.00027606,0.00000873186],"category_scores_gemma":[0.0001960985,0.000306267,0.0001208548,0.000369033,0.0000623681,0.0006527805,0.001329103,0.0002843202,0.000001929284],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000664002,"about_ca_system_score_gemma":0.0007223474,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005960039,"about_ca_topic_score_gemma":0.00001248324,"domain_scores_codex":[0.9983852,0.0001366082,0.0002104828,0.0008439531,0.0001627111,0.0002610252],"domain_scores_gemma":[0.9978157,0.0001121308,0.000379732,0.0008123722,0.0007746838,0.0001053481],"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":[0.000007649695,0.0000148788,0.00004737743,0.00005773851,0.00005120805,0.000008026879,0.0002200193,0.8719923,0.000001116615,0.1269061,0.00001643148,0.0006771166],"study_design_scores_gemma":[0.0005564871,0.00004220908,0.00003512131,0.00009453867,0.00009874719,0.000002229702,0.00006507335,0.9893267,0.00000980292,0.009473529,0.000009132325,0.0002864534],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.003084703,0.0000511023,0.9947155,0.0002232617,0.0003557781,0.0006698938,0.000009969463,0.0001484375,0.0007413602],"genre_scores_gemma":[0.9092041,0.0002246643,0.08896544,0.00008286817,0.00008689213,0.000004634834,0.0003153012,0.00002205293,0.001094102],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9061193,"threshold_uncertainty_score":0.999939,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08714489944938392,"score_gpt":0.2331941443887187,"score_spread":0.1460492449393348,"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."}}