{"id":"W4405254983","doi":"10.48550/arxiv.2412.07165","title":"A Method for Evaluating Hyperparameter Sensitivity in Reinforcement Learning","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Evolutionary Algorithms and Applications","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Alliance de recherche numérique du Canada; Alberta Machine Intelligence Institute; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Hyperparameter; Sensitivity (control systems); Reinforcement learning; Reinforcement; Machine learning; Artificial intelligence; Psychology; Computer science; Social psychology; Engineering","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02192893,0.002143864,0.001602202,0.004693119,0.001197032,0.002715854,0.002882648,0.003137233,0.002592155],"category_scores_gemma":[0.1371896,0.0008915514,0.001394701,0.003072992,0.003573734,0.003166545,0.003511116,0.005695172,0.000552415],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001819611,"about_ca_system_score_gemma":0.001735803,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002336551,"about_ca_topic_score_gemma":0.001545574,"domain_scores_codex":[0.9838774,0.007989405,0.001095814,0.002434035,0.004183775,0.0004196805],"domain_scores_gemma":[0.9052563,0.07227198,0.007206001,0.01001137,0.004597298,0.0006570737],"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":[0.0004948049,0.0004070871,0.01644486,0.0005373135,0.0006936666,0.0001783921,0.0006168215,0.7368044,0.008024438,0.03862675,0.00292354,0.194248],"study_design_scores_gemma":[0.00008407124,0.0004224647,0.004560263,0.0001371884,0.00008438885,0.0002645428,0.00009730463,0.9402254,0.007977146,0.04330279,0.002724308,0.0001201877],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02545139,0.0005805448,0.9694531,0.0002076936,0.00009760784,0.0004151571,0.0001911687,0.0009739074,0.002629424],"genre_scores_gemma":[0.4582496,0.0003607423,0.5364583,0.000375453,0.00009835267,0.00214976,0.0005262076,0.0005198051,0.001261831],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02192893,"threshold_uncertainty_score":0.1159726,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1114899556047503,"score_gpt":0.2707874725826567,"score_spread":0.1592975169779064,"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."}}