{"id":"W3043761458","doi":"","title":"Beyond Prioritized Replay: Sampling States in Model-Based RL via Simulated Priorities","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Advanced Bandit Algorithms Research","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Sampling (signal processing); Key (lock); Benchmark (surveying); Equivalence (formal languages); Sample (material); Ideal (ethics); Mathematical optimization; Sampling bias; Sample size determination; Mathematics; Statistics; Telecommunications","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":[0.002422416,0.0006337453,0.001213588,0.001644426,0.0003110643,0.0006161941,0.002335479,0.0006604522,0.0003231331],"category_scores_gemma":[0.002101835,0.0006864729,0.0004626615,0.002758403,0.0004658705,0.000748741,0.002323248,0.001853605,0.00008176857],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008747295,"about_ca_system_score_gemma":0.001465468,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004072758,"about_ca_topic_score_gemma":0.0005462212,"domain_scores_codex":[0.9934272,0.0006972214,0.001045783,0.002789099,0.001039265,0.001001412],"domain_scores_gemma":[0.9920746,0.00302475,0.0005649668,0.002425666,0.001532859,0.000377121],"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":[0.0002779061,0.0001223727,0.003966434,0.00005672121,0.00004562673,0.000925458,0.0003842675,0.9924144,0.00008970419,0.0003489099,0.00003827983,0.001329966],"study_design_scores_gemma":[0.001466452,0.00003637769,0.0005640098,0.0001514871,0.00002857894,0.000002414808,0.0007251087,0.8177481,0.0003901781,0.178196,0.0001055708,0.000585671],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4608836,0.0001900997,0.537587,0.00007154291,0.0003175289,0.0004888928,0.00008984642,0.000123629,0.0002478199],"genre_scores_gemma":[0.988839,0.0002215166,0.007942921,0.0001028527,0.0000765992,0.000002653579,0.0001474142,0.00006763388,0.002599365],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5296441,"threshold_uncertainty_score":0.9995586,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2194841626866739,"score_gpt":0.3263061400790038,"score_spread":0.1068219773923299,"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."}}