{"id":"W7124312491","doi":"10.65109/octa6133","title":"Targeted Search Control in AlphaZero for Effective Policy Improvement","year":2023,"lang":"","type":"article","venue":"","topic":"Artificial Intelligence in Games","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Reinforcement learning; Sample (material); Limiting; Control (management); Value (mathematics); State (computer science); Train; Function (biology)","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","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002555195,0.0004226173,0.0005677376,0.001149432,0.0002823004,0.0004086714,0.001511608,0.0002251944,0.0001205323],"category_scores_gemma":[0.0009300223,0.0004110147,0.0002728393,0.00369423,0.0002606674,0.0005937385,0.0006365387,0.0003634581,0.001257721],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005770975,"about_ca_system_score_gemma":0.0007084569,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00351138,"about_ca_topic_score_gemma":0.0005094403,"domain_scores_codex":[0.9950825,0.0003000452,0.0009207944,0.001225729,0.0006628886,0.001808062],"domain_scores_gemma":[0.9962775,0.001996422,0.0001328246,0.0008507835,0.0004424934,0.0002999845],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002049785,0.0002775126,0.0008709827,0.0001073172,0.0001099029,0.00003324606,0.006457514,0.01006593,0.02896791,0.06749389,0.002638851,0.882772],"study_design_scores_gemma":[0.0008581607,0.001343975,0.001856926,0.00005259859,0.000009621046,0.000001030357,0.0006003485,0.8469831,0.1292552,0.01726385,0.001370269,0.0004048575],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05478829,0.0001270443,0.9193757,0.01561501,0.001539196,0.006942523,0.00005556239,0.0003768098,0.001179934],"genre_scores_gemma":[0.9910416,0.0000584104,0.00231625,0.001811378,0.0004977292,0.001081777,0.000005403366,0.00004141423,0.003146046],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9362533,"threshold_uncertainty_score":0.9998342,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0306515821686123,"score_gpt":0.3479757694916265,"score_spread":0.3173241873230142,"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."}}