{"id":"W7124291073","doi":"10.65109/yptr7088","title":"MaDi: Learning to Mask Distractions for Generalization in Visual Deep Reinforcement Learning","year":2024,"lang":"","type":"article","venue":"","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Generalization; Reinforcement learning; Focus (optics); Artificial neural network; Contrast (vision); Deep learning; Control (management)","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001414291,0.001514061,0.0009226219,0.0003525171,0.0003387376,0.0006563731,0.00232523,0.001128184,0.001952123],"category_scores_gemma":[0.003657062,0.0004735457,0.0005018204,0.0002023935,0.0009337825,0.001093805,0.001581243,0.002338645,0.0003980763],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001215563,"about_ca_system_score_gemma":0.001303212,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004509468,"about_ca_topic_score_gemma":0.005379159,"domain_scores_codex":[0.9996349,0.00007384043,0.00001883427,0.0001092188,0.00008953475,0.00007369775],"domain_scores_gemma":[0.9991866,0.0003524935,0.0001139119,0.0001615725,0.00009984148,0.00008569844],"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":[0.0002971031,0.0002814781,0.002263916,0.0001492876,0.0001100908,0.00009155474,0.0001450249,0.7177654,0.01355495,0.006777155,0.005683966,0.2528801],"study_design_scores_gemma":[0.0000249972,0.00007925009,0.0001352711,0.000008140531,0.000008640086,0.00001593796,0.000005783754,0.9939716,0.003008298,0.002265453,0.0004711931,0.000005339669],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07641535,0.0007867812,0.9101005,0.000465916,0.0001335781,0.0001551196,0.0001371052,0.008098305,0.003707411],"genre_scores_gemma":[0.7983482,0.0001556473,0.1964888,0.0004712637,0.000048,0.0002236517,0.0003063517,0.0003089833,0.003649091],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004509468,"threshold_uncertainty_score":0.008966446,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02183426322905901,"score_gpt":0.3108022284253805,"score_spread":0.2889679651963215,"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."}}