{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002963952,0.001570163,0.001510593,0.0007613683,0.0004571015,0.001025413,0.003407438,0.001786315,0.003992889],"category_scores_gemma":[0.008099501,0.0006680427,0.0007215178,0.0004839851,0.001835933,0.001292611,0.002145733,0.002902744,0.0006958628],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001509508,"about_ca_system_score_gemma":0.002170761,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004296772,"about_ca_topic_score_gemma":0.005412303,"domain_scores_codex":[0.9990659,0.0003692004,0.00004480981,0.0001926194,0.0002061002,0.0001212872],"domain_scores_gemma":[0.9974492,0.001518527,0.0002429954,0.0002303943,0.0003729102,0.0001858608],"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":[0.0001487363,0.0002126072,0.001394576,0.0001769784,0.0000909873,0.0001012446,0.0001325193,0.8540884,0.002760768,0.02196579,0.003344916,0.1155826],"study_design_scores_gemma":[0.0000211617,0.00004608157,0.00005181342,0.000009630488,0.000005082447,0.00001115088,0.000003686999,0.9927145,0.000338225,0.006336604,0.0004561569,0.000005967492],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.007374036,0.0002739302,0.9896254,0.0002491584,0.00003898202,0.0001131796,0.00006554285,0.0008719119,0.001387902],"genre_scores_gemma":[0.6026455,0.0002694089,0.3917251,0.000608296,0.0001095753,0.0007638416,0.0002888082,0.0003269286,0.00326264],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004296772,"threshold_uncertainty_score":0.01567501,"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."}}