{"id":"W7015655229","doi":"","title":"Towards alignment of Reinforcement Learning agents; for consideration of safety, robustness and fairness.","year":2024,"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":"Robustness (evolution); Reinforcement learning; Control theory (sociology); 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":[0.002015038,0.0006547939,0.0006349349,0.0003402467,0.0006172977,0.001263657,0.00104447,0.001130773,0.008618779],"category_scores_gemma":[0.009103605,0.0004031367,0.000652857,0.0004384186,0.001811142,0.00231807,0.002847794,0.003170828,0.001376128],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001191912,"about_ca_system_score_gemma":0.00121079,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001135327,"about_ca_topic_score_gemma":0.0009652679,"domain_scores_codex":[0.999042,0.0004065188,0.00004086701,0.0001975,0.0002238617,0.00008924012],"domain_scores_gemma":[0.998268,0.0009079602,0.0002162453,0.0002206603,0.0001941133,0.0001930025],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00008896382,0.00005627678,0.0005564871,0.0001030073,0.00004784259,0.00006520363,0.0002913494,0.1161632,0.001561,0.8072919,0.007751024,0.06602363],"study_design_scores_gemma":[0.0000331431,0.00005620916,0.0002458524,0.0000519219,0.00001580196,0.00003420164,0.00005666779,0.3551507,0.0009462617,0.6243763,0.01901994,0.00001295181],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01021241,0.0011592,0.9570061,0.001869164,0.0002446482,0.00007260001,0.00005467571,0.0002787965,0.02910244],"genre_scores_gemma":[0.5126585,0.001479494,0.4473557,0.0008762159,0.0004217574,0.0003551937,0.0002501467,0.0004639823,0.03613897],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008618779,"threshold_uncertainty_score":0.02883267,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02355615140994175,"score_gpt":0.2650591302153393,"score_spread":0.2415029788053975,"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."}}