{"id":"W2916603395","doi":"10.24963/ijcai.2019/386","title":"Perturbed-History Exploration in Stochastic Multi-Armed Bandits","year":2019,"lang":"en","type":"preprint","venue":"","topic":"Advanced Bandit Algorithms Research","field":"Decision Sciences","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Regret; Bernoulli's principle; Offset (computer science); Argument (complex analysis); Computer science; Optimism; Mathematical optimization; Mathematics; Psychology; Machine learning; Engineering; Social psychology","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":["insufficient_payload"],"category_scores_codex":[0.003012296,0.0004165772,0.0007927764,0.001552461,0.00005958096,0.0002188629,0.001868087,0.0005114861,0.002890768],"category_scores_gemma":[0.008142646,0.0003217224,0.0002102414,0.0005004869,0.000185589,0.0007500534,0.001642358,0.001472301,0.003618156],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001459594,"about_ca_system_score_gemma":0.001008755,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001915798,"about_ca_topic_score_gemma":0.0005539819,"domain_scores_codex":[0.9930082,0.0004794098,0.001156381,0.001631684,0.003121385,0.0006029534],"domain_scores_gemma":[0.9945634,0.002333641,0.0004124111,0.001773136,0.0007124969,0.0002048967],"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":[0.0001335631,0.0003199152,0.0003954657,0.00003956392,0.00002750982,0.00006387584,0.00408062,0.9327971,0.0001304415,0.0001104403,0.03478238,0.02711909],"study_design_scores_gemma":[0.001648637,0.00008464836,0.00390616,0.0001343279,0.000008127839,0.000005250315,0.001509555,0.9592872,0.00009648468,0.01745185,0.01508492,0.0007828798],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.003095024,0.001049805,0.987567,0.0005581289,0.004094683,0.001465227,0.00002775802,0.0001092141,0.002033142],"genre_scores_gemma":[0.8888991,0.0001061202,0.01172835,0.0002805285,0.0004398506,0.000452391,0.000101426,0.00008790088,0.09790435],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9758387,"threshold_uncertainty_score":0.9999235,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4035207061190713,"score_gpt":0.4707663724538747,"score_spread":0.06724566633480344,"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."}}