{"id":"W2977481643","doi":"10.1609/aaai.v35i12.17276","title":"Improving Sample Efficiency in Model-Free Reinforcement Learning from Images","year":2021,"lang":"en","type":"preprint","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":121,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Reinforcement learning; Robustness (evolution); Computer science; Artificial intelligence; Stability (learning theory); Machine learning; Encoder; Representation (politics); Noise (video); Code (set theory); Image (mathematics)","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":[0.003099155,0.001084658,0.001578146,0.000487317,0.0005415608,0.001027408,0.001959354,0.001416929,0.00241057],"category_scores_gemma":[0.01484934,0.0007995028,0.0005556846,0.0003333554,0.001785584,0.00218686,0.001945493,0.0022878,0.0004862301],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00124865,"about_ca_system_score_gemma":0.001641659,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004899444,"about_ca_topic_score_gemma":0.004968627,"domain_scores_codex":[0.9991339,0.0003630647,0.00004180688,0.0001845011,0.0001717569,0.0001050624],"domain_scores_gemma":[0.9942808,0.004170848,0.0003821589,0.00065927,0.0003222139,0.000184822],"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":[0.0002275586,0.0001296509,0.001024837,0.00008547858,0.00003937187,0.00007330465,0.000102177,0.9321941,0.002344172,0.01676286,0.0008702454,0.04614608],"study_design_scores_gemma":[0.00001313598,0.00002218607,0.00004434441,0.000004713864,0.000002714956,0.000006787918,0.000003214012,0.9947321,0.0004665226,0.004617885,0.00008383371,0.000002548156],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04095241,0.0001798439,0.9556742,0.0003145358,0.00002435989,0.00005164371,0.00003249415,0.001066449,0.001704109],"genre_scores_gemma":[0.8612784,0.00009830592,0.1359755,0.0002162296,0.00002824301,0.0001476574,0.0001081945,0.000244299,0.001903166],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004899444,"threshold_uncertainty_score":0.01639009,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07127289501225326,"score_gpt":0.2836056152858242,"score_spread":0.2123327202735709,"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."}}