{"id":"W4285600734","doi":"10.24963/ijcai.2022/478","title":"CCLF: A Contrastive-Curiosity-Driven Learning Framework for Sample-Efficient Reinforcement Learning","year":2022,"lang":"en","type":"article","venue":"Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"BC Research (Canada)","funders":"Nanyang Technological University; National Research Foundation Singapore; National Research Foundation","keywords":"Computer science; Reinforcement learning; Curiosity; Artificial intelligence; Encoder; Exploit; Sample (material); Machine learning; Code (set theory); Encoding (memory); Representation (politics)","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","sts"],"consensus_categories":[],"category_scores_codex":[0.001251875,0.0003715901,0.0004026537,0.0003437517,0.001419109,0.0004802821,0.003123904,0.00009863816,0.0003242398],"category_scores_gemma":[0.00382232,0.0003351152,0.0003299236,0.0006086847,0.0002314833,0.0003124301,0.001788158,0.001414028,0.0000368409],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004472782,"about_ca_system_score_gemma":0.0001631855,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004863316,"about_ca_topic_score_gemma":0.000001979891,"domain_scores_codex":[0.9959245,0.0000632169,0.001049661,0.0007488555,0.001628684,0.0005851319],"domain_scores_gemma":[0.99624,0.0008829917,0.001294205,0.0003014569,0.001154,0.000127385],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00008984858,0.00007589168,0.0002860164,0.00002153708,0.00004622113,4.413612e-7,0.001702382,0.4939584,0.0008564221,0.5012636,0.0000322992,0.001666921],"study_design_scores_gemma":[0.0001115837,0.0007307802,0.0001315525,0.0002704457,0.00002128785,0.000008293553,0.001675023,0.9336187,0.01537294,0.04546282,0.002245376,0.0003512482],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.008247299,0.000008993653,0.9794016,0.006298661,0.001607997,0.0009196901,0.000009448588,0.0001949413,0.003311334],"genre_scores_gemma":[0.9763002,0.0000241019,0.0223253,0.0003808046,0.0001466768,0.0002352311,0.000009193174,0.00003064173,0.000547786],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.968053,"threshold_uncertainty_score":0.9999101,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06430454282604245,"score_gpt":0.2937313023092388,"score_spread":0.2294267594831964,"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."}}