{"id":"W3090729135","doi":"10.1109/ijcnn48605.2020.9207648","title":"Improved Policy Extraction via Online Q-Value Distillation","year":2020,"lang":"en","type":"article","venue":"","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Distillation; Artificial neural network; Decision tree; Function (biology); Artificial intelligence; Monte Carlo tree search; Machine learning; Value (mathematics); Monte Carlo method; Algorithm; Mathematics","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":[0.001907998,0.001057528,0.001525667,0.0006493921,0.0007101988,0.00129046,0.00169045,0.001783197,0.005101834],"category_scores_gemma":[0.009500662,0.0007316516,0.0007817817,0.0008131991,0.001192861,0.002656367,0.002053734,0.00283473,0.001170231],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001042363,"about_ca_system_score_gemma":0.003056988,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004996561,"about_ca_topic_score_gemma":0.006760841,"domain_scores_codex":[0.9989169,0.0003300488,0.0000544208,0.0002244204,0.0003281706,0.00014611],"domain_scores_gemma":[0.9967768,0.002214229,0.0001895107,0.0004689687,0.0002510187,0.00009941042],"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":[0.0002821444,0.000167999,0.0008607068,0.0001525813,0.00006216671,0.0001540555,0.0001107369,0.7117016,0.003099309,0.0452254,0.003826084,0.2343571],"study_design_scores_gemma":[0.00002256591,0.00001570532,0.00003492671,0.000008049939,0.000004517808,0.00001095974,0.00000378713,0.985293,0.0008473165,0.01323628,0.0005176191,0.000005374411],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01898563,0.0003588078,0.9753161,0.0003400151,0.00007778344,0.00007795692,0.0001529516,0.002333221,0.002357399],"genre_scores_gemma":[0.602853,0.0002046044,0.3927758,0.0004064481,0.00006725624,0.0001861655,0.0004475384,0.0004706468,0.002588549],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005101834,"threshold_uncertainty_score":0.01706731,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03551764141124086,"score_gpt":0.3154533688888263,"score_spread":0.2799357274775854,"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."}}