{"id":"W2963106920","doi":"10.1177/0278364920910802","title":"Improving user specifications for robot behavior through active preference learning: Framework and evaluation","year":2020,"lang":"en","type":"article","venue":"The International Journal of Robotics Research","topic":"Robot Manipulation and Learning","field":"Engineering","cited_by":28,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Task (project management); Robot; Process (computing); Preference; Preference elicitation; Preference learning; Task analysis; Active learning (machine learning)","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01715295,0.001771762,0.001086397,0.001145066,0.0004927833,0.001685006,0.003219877,0.002394733,0.002483722],"category_scores_gemma":[0.07422473,0.0005393475,0.0008744156,0.0008203389,0.001278377,0.002655251,0.002072214,0.002354521,0.000627577],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001798952,"about_ca_system_score_gemma":0.002337392,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01051645,"about_ca_topic_score_gemma":0.008546645,"domain_scores_codex":[0.9858472,0.00926239,0.0007107527,0.001131558,0.002544705,0.0005034558],"domain_scores_gemma":[0.9009092,0.07965502,0.00329522,0.006193073,0.008475033,0.00147248],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.006499477,0.01259194,0.03358021,0.002505915,0.0004250253,0.0002525689,0.002605521,0.2779555,0.0179511,0.007348234,0.004317844,0.6339667],"study_design_scores_gemma":[0.0004569029,0.002420413,0.003712517,0.00006835564,0.00009337549,0.00007683187,0.000316472,0.9822405,0.00724106,0.002357653,0.0009364735,0.00007940659],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4138249,0.0008382772,0.5738992,0.0004025681,0.00004299671,0.002048013,0.0004337611,0.005017057,0.003493295],"genre_scores_gemma":[0.8133045,0.0001724727,0.1838917,0.0001391777,0.00001287464,0.0009975639,0.0005957956,0.0001526683,0.0007331347],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01715295,"threshold_uncertainty_score":0.09071457,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4275317402355947,"score_gpt":0.4292571667206051,"score_spread":0.001725426485010395,"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."}}