{"id":"W4367016230","doi":"10.1109/tse.2023.3269804","title":"A Search-Based Testing Approach for Deep Reinforcement Learning Agents","year":2023,"lang":"en","type":"article","venue":"IEEE Transactions on Software Engineering","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":46,"is_retracted":false,"has_abstract":true,"ca_institutions":"Cisco Systems (Canada); École de Technologie Supérieure; University of Ottawa","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Reinforcement learning; Artificial intelligence; Machine learning; Software testing; Software engineering; Programming language; Software","routes":{"ca_aff":true,"ca_fund":true,"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"],"consensus_categories":[],"category_scores_codex":[0.0004131361,0.0002458966,0.0001909584,0.0004949536,0.0003766184,0.0001668999,0.0005319753,0.00008985886,0.000008817609],"category_scores_gemma":[0.0001670887,0.0002759447,0.0001393434,0.001272861,0.00001857685,0.0002812286,0.000008164444,0.0004429373,0.00006013378],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001552086,"about_ca_system_score_gemma":0.00006328034,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006494386,"about_ca_topic_score_gemma":1.312699e-7,"domain_scores_codex":[0.9981533,0.00002727279,0.0003108435,0.0004230914,0.0004671755,0.0006182844],"domain_scores_gemma":[0.9986612,0.0006063294,0.00005963717,0.000419785,0.0001115752,0.0001414871],"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":[0.000005305322,0.00001604539,0.00006023975,0.0001308153,0.00003367685,0.000004336891,0.0002360686,0.9935234,0.000180938,0.00005137323,0.00002578865,0.005732055],"study_design_scores_gemma":[0.0004786784,0.0001925355,0.00009300671,0.00006190662,0.00001292147,0.000003133624,0.00002304759,0.9951137,0.003417718,0.000001475469,0.0003245971,0.000277334],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0006383811,0.00000454791,0.9957328,0.0000219518,0.0004634642,0.0004573712,0.000001023926,0.002620914,0.00005949046],"genre_scores_gemma":[0.5927373,0.00000244421,0.406452,0.00004813849,0.00003526558,0.0001880706,0.00001149806,0.00004787507,0.0004774302],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.592099,"threshold_uncertainty_score":0.9999692,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04544628432889011,"score_gpt":0.2572917273298779,"score_spread":0.2118454430009878,"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."}}