{"id":"W4414943628","doi":"10.48550/arxiv.2507.10797","title":"Multi-Armed Sampling Problem and the End of Exploration","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Machine Learning and Algorithms","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Institut de Valorisation des Données; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Regret; Reinforcement learning; Sampling (signal processing); Thompson sampling; Convergence (economics); Context (archaeology); Sample (material); Contrast (vision)","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006497119,0.0001609164,0.0002603852,0.00008146255,0.0001225626,0.00009311201,0.0006732033,0.00009927767,0.000003744231],"category_scores_gemma":[0.0001382412,0.0001123934,0.00007437666,0.0001467559,0.0001138035,0.0001231435,0.001235064,0.0005132556,0.000005097979],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001445815,"about_ca_system_score_gemma":0.00009827421,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002876102,"about_ca_topic_score_gemma":0.00001456171,"domain_scores_codex":[0.9988012,0.000176982,0.0002871544,0.0004249544,0.0001619922,0.0001477344],"domain_scores_gemma":[0.9988251,0.0002409928,0.0002351181,0.0005817966,0.00008438267,0.00003263527],"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":[0.0001124258,0.0004970821,0.2564651,0.00274038,0.0006198359,0.00001665035,0.05174643,0.07716268,0.002040796,0.06339347,0.0004320911,0.544773],"study_design_scores_gemma":[0.002439568,0.00005765022,0.06397236,0.0008007768,0.00007119717,0.000005102897,0.0001761947,0.9110633,0.001257946,0.01696431,0.002651443,0.0005401605],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09160861,0.000559852,0.9022043,0.003842695,0.0005673967,0.0004278689,0.000005428012,0.0001362678,0.000647537],"genre_scores_gemma":[0.8440783,0.0001907889,0.1544107,0.0001768087,0.0001179131,0.00007961924,0.00001329764,0.00000835982,0.0009241824],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8339006,"threshold_uncertainty_score":0.4583271,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06852553140865048,"score_gpt":0.3101442841313991,"score_spread":0.2416187527227486,"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."}}