{"id":"W3172765438","doi":"10.48550/arxiv.2012.13658","title":"Locally Persistent Exploration in Continuous Control Tasks with Sparse\\n Rewards","year":2020,"lang":"","type":"preprint","venue":"arXiv (Cornell University)","topic":"Data Visualization and Analytics","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Reinforcement learning; Computer science; State space; Trajectory; State (computer science); Action (physics); Task (project management); Space (punctuation); Artificial intelligence; Algorithm; 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":[0.0005325823,0.0007179016,0.0009856609,0.0005123951,0.0002394097,0.0005640834,0.002292313,0.0004022035,0.0001081154],"category_scores_gemma":[0.0001076643,0.0008075252,0.0003816168,0.002120102,0.0003519459,0.001498937,0.001196588,0.0009225446,0.0002955232],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000504338,"about_ca_system_score_gemma":0.0009167252,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002754901,"about_ca_topic_score_gemma":0.0003709485,"domain_scores_codex":[0.9953649,0.0005291431,0.0006869728,0.002361765,0.0003620014,0.0006951992],"domain_scores_gemma":[0.9964488,0.000111148,0.0007509075,0.001514452,0.000623366,0.0005513622],"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":[0.0005612385,0.0005399175,0.006943741,0.0001913178,0.0004169696,0.002073552,0.002192409,0.8483012,0.00003340445,0.1369033,0.0006349,0.001208097],"study_design_scores_gemma":[0.003887294,0.0006527713,0.0006294711,0.000425495,0.0003319925,0.00001140747,0.001649099,0.9863808,0.00002960705,0.0021843,0.002858458,0.0009593382],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01495208,0.00006394666,0.9798498,0.001384605,0.0004378071,0.0008811036,0.00009851826,0.000182118,0.002150082],"genre_scores_gemma":[0.9959486,0.0004859115,0.0006975385,0.001060137,0.0001093216,0.000002463402,0.0001554077,0.00004236868,0.001498246],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9809965,"threshold_uncertainty_score":0.9994376,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09598491008680228,"score_gpt":0.207414975584561,"score_spread":0.1114300654977588,"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."}}