{"id":"W4286910437","doi":"10.48550/arxiv.2110.02355","title":"Robustness and sample complexity of model-based MARL for general-sum Markov games","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Ministère de la Défense Nationale","keywords":"Markov chain; Mathematics; Markov kernel; Markov process; Markov perfect equilibrium; Discrete mathematics; Markov model; Applied mathematics; Combinatorics; Mathematical optimization; Variable-order Markov model; Nash equilibrium; Statistics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0002722447,0.0003084441,0.0004876041,0.0002079367,0.0001358937,0.0001448781,0.001286118,0.000237175,0.00001755774],"category_scores_gemma":[0.0001092372,0.0003773474,0.0002242924,0.0003101383,0.0002830808,0.0002395377,0.001753168,0.0003232897,5.193042e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001214062,"about_ca_system_score_gemma":0.0003579812,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000136748,"about_ca_topic_score_gemma":0.00003080608,"domain_scores_codex":[0.9981956,0.0001139758,0.0002764248,0.0009435415,0.0001253137,0.0003451744],"domain_scores_gemma":[0.9976226,0.0003164615,0.0003752711,0.001194652,0.0003554029,0.0001356119],"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":[0.0000265612,0.00003849902,0.001369851,0.0003527713,0.00006359435,0.00001384015,0.00007560095,0.9633426,0.00002801363,0.03440688,0.00008960454,0.0001921842],"study_design_scores_gemma":[0.0006108685,0.00004979336,0.0005436922,0.0001024369,0.00005744677,9.099177e-7,0.00003363084,0.9924962,0.0002358572,0.005465009,0.00003736452,0.0003668208],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1073999,0.00003483954,0.8915175,0.00009211755,0.0002434581,0.0003634434,0.0000364543,0.00009399506,0.0002183234],"genre_scores_gemma":[0.6982971,0.00004643606,0.3008869,0.0000614678,0.00002307618,0.000001693734,0.00005839523,0.00001673822,0.0006081713],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5908973,"threshold_uncertainty_score":0.9998679,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.12315007179665,"score_gpt":0.2188418402903258,"score_spread":0.0956917684936758,"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."}}