{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00009771966,0.0001105858,0.000102935,0.00007187396,0.0001050167,0.0001267585,0.000427796,0.00005204618,0.0000508295],"category_scores_gemma":[0.0002120597,0.0001032813,0.00005095169,0.0006937854,0.00002383784,0.0008544032,0.0001346172,0.0001137121,0.0002483214],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007166553,"about_ca_system_score_gemma":0.00008070082,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000822856,"about_ca_topic_score_gemma":0.00008607085,"domain_scores_codex":[0.9989405,0.00004015702,0.0002435478,0.0003566267,0.0001789491,0.0002402785],"domain_scores_gemma":[0.9993094,0.00005957008,0.00007853638,0.0003019454,0.00009079315,0.0001597093],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00002264483,0.0001481177,0.0002316917,0.00002371506,0.00001610048,0.00001415092,0.001750133,0.003811065,0.2143587,0.419297,0.0007486106,0.3595781],"study_design_scores_gemma":[0.00004979276,0.00009990095,0.0006680608,0.00000235147,0.000002160089,0.000005370165,0.00007130599,0.9295495,0.0591334,0.005803814,0.004479447,0.0001349425],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01012642,0.00001461484,0.9660599,0.02020512,0.0001663893,0.0001503846,0.000001633695,0.0003813324,0.002894213],"genre_scores_gemma":[0.9484518,0.00000836716,0.04801906,0.002750889,0.0004545264,0.000004544944,0.00000654753,0.000009000972,0.000295338],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9383253,"threshold_uncertainty_score":0.4211689,"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."}}