{"id":"W4313483332","doi":"10.48550/arxiv.2301.00051","title":"Learning from Guided Play: Improving Exploration for Adversarial Imitation Learning with Simple Auxiliary Tasks","year":2022,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Rehabilitation Institute; University of Alberta","funders":"","keywords":"Leverage (statistics); Computer science; Artificial intelligence; Machine learning; Reinforcement learning; Task (project management); Adversarial system; Imitation; 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":[0.0005692893,0.0004295742,0.000405243,0.0003519827,0.0009036537,0.0003524054,0.001389172,0.0002658718,0.00006903733],"category_scores_gemma":[0.0003534827,0.0005233329,0.0002047845,0.0005246932,0.00006711616,0.001370358,0.001969068,0.00146035,0.00002057979],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005889594,"about_ca_system_score_gemma":0.0004660928,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005841276,"about_ca_topic_score_gemma":0.00003510986,"domain_scores_codex":[0.9970431,0.0003958566,0.000381983,0.001401658,0.0002835393,0.0004938063],"domain_scores_gemma":[0.9972983,0.0005533829,0.0009166987,0.0008216052,0.0002762268,0.0001337604],"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":[0.000124203,0.00002058283,0.002578744,0.00005359841,0.0001212638,0.00005422302,0.001819808,0.988102,0.0001245653,0.005990316,0.0001204725,0.0008902373],"study_design_scores_gemma":[0.001364463,0.0004756195,0.0002922164,0.00004857744,0.000136514,0.000001256447,0.001429009,0.9908108,0.0000937732,0.002882716,0.001864151,0.0006009066],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07804786,0.00001121307,0.9192711,0.0000681922,0.0007009891,0.0007566994,0.000008561516,0.0005467762,0.0005886717],"genre_scores_gemma":[0.9784813,0.00003056609,0.01838433,0.00005049799,0.0001982343,0.00001635633,0.0008986435,0.00005801663,0.001882036],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9008867,"threshold_uncertainty_score":0.9997218,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09622267453762151,"score_gpt":0.2116297504875747,"score_spread":0.1154070759499532,"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."}}