{"id":"W7015986514","doi":"","title":"Towards better generalization capabilities of reinforcement learning agents via self-supervision","year":2023,"lang":"en","type":"dissertation","venue":"eScholarship@McGill (McGill)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Generalization; Reinforcement learning; Control (management); Feature (linguistics); Stability (learning theory)","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":[0.003042219,0.0008834894,0.0009760351,0.0003474503,0.0003359283,0.0007967641,0.001374531,0.0009966566,0.001137879],"category_scores_gemma":[0.009958453,0.0005475017,0.0006918507,0.0002281949,0.00150696,0.001983671,0.002102312,0.002480446,0.0002955691],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009292933,"about_ca_system_score_gemma":0.0009895783,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003379563,"about_ca_topic_score_gemma":0.002755624,"domain_scores_codex":[0.9991912,0.0003169196,0.00004792004,0.000194742,0.000162019,0.00008725786],"domain_scores_gemma":[0.9952538,0.00260273,0.000473373,0.0009973111,0.0004741657,0.0001985658],"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":[0.00008489422,0.0000970927,0.00105238,0.00004814094,0.00005657634,0.00005312413,0.0001656685,0.9223219,0.00421379,0.02285152,0.0006357517,0.04841915],"study_design_scores_gemma":[0.00000614577,0.00002434239,0.0000434273,0.000002385185,0.000002028632,0.000003834849,0.00000298236,0.9948425,0.0003121564,0.004679281,0.00007882273,0.000002123368],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07170885,0.000131368,0.9251217,0.0004006594,0.00002265422,0.00006267463,0.00002350487,0.0007756297,0.001752904],"genre_scores_gemma":[0.8805702,0.00008998442,0.1172923,0.0001684865,0.00002660599,0.0001084256,0.00005121899,0.00007223614,0.001620534],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003379563,"threshold_uncertainty_score":0.01608902,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01924933637974323,"score_gpt":0.2519997002253784,"score_spread":0.2327503638456352,"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."}}