{"id":"W2890047470","doi":"10.48550/arxiv.1811.09013","title":"An Off-policy Policy Gradient Theorem Using Emphatic Weightings","year":2018,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Unicode; Counterexample; Reinforcement learning; Computer science; Key (lock); Mathematical economics; Mathematics; Mathematical optimization; Artificial intelligence; Discrete mathematics; Computer security","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002704705,0.001012218,0.0008999616,0.0004352611,0.0004303576,0.001107966,0.001253444,0.001344935,0.004532873],"category_scores_gemma":[0.01148087,0.0005729345,0.0005321449,0.0003255593,0.002046587,0.002243493,0.00193892,0.002668649,0.000783656],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009833711,"about_ca_system_score_gemma":0.001443979,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001699596,"about_ca_topic_score_gemma":0.001483857,"domain_scores_codex":[0.9991378,0.0003305099,0.0000398208,0.0001639793,0.0002519081,0.00007603981],"domain_scores_gemma":[0.9978979,0.001277267,0.0001760396,0.000240891,0.0003058699,0.0001020219],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001226249,0.00006136678,0.0007815849,0.0001379147,0.00004502029,0.0001177131,0.000154199,0.4976787,0.003944141,0.4278703,0.003565352,0.06552111],"study_design_scores_gemma":[0.00001548888,0.00003965915,0.00008173227,0.00001555722,0.000005676402,0.00003115166,0.000006820874,0.9384211,0.0008531216,0.05906294,0.00145883,0.000007990525],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.005860225,0.00008148948,0.9894386,0.0002858247,0.00004882177,0.00003119267,0.00001821023,0.0001520527,0.004083597],"genre_scores_gemma":[0.5574764,0.0003714469,0.4279793,0.0006569764,0.0001359516,0.0002793777,0.0001176743,0.0003758803,0.01260697],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004532873,"threshold_uncertainty_score":0.01516396,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06837137357235425,"score_gpt":0.2310214684100404,"score_spread":0.1626500948376862,"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."}}