{"id":"W4367060663","doi":"10.48550/arxiv.2304.12567","title":"Proto-Value Networks: Scaling Representation Learning with Auxiliary Tasks","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Reinforcement learning; Computer science; Successor cardinal; Representation (politics); Value (mathematics); Artificial intelligence; Function (biology); Carry (investment); Feature learning; Object (grammar); Machine learning; Mathematics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0005516869,0.0004036499,0.000383206,0.0003632519,0.0003753808,0.0003711994,0.001833311,0.0003304079,0.00001177037],"category_scores_gemma":[0.0001236748,0.000442894,0.0001701601,0.001242335,0.0001306412,0.0005945344,0.002614018,0.001537825,0.0001374625],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002612071,"about_ca_system_score_gemma":0.0002245344,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002404269,"about_ca_topic_score_gemma":0.00000980268,"domain_scores_codex":[0.9970387,0.000354024,0.0003157329,0.001467707,0.0002581691,0.0005656871],"domain_scores_gemma":[0.9974083,0.0002493538,0.0005474925,0.001410426,0.0002095543,0.0001748332],"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":[0.00003024725,0.00001583441,0.01504152,0.00007245222,0.0001086209,0.0003567452,0.0002758131,0.9673922,0.000003724359,0.01632145,0.0001114155,0.0002699944],"study_design_scores_gemma":[0.0004102486,0.00009699726,0.00232359,0.0003327671,0.00006619595,0.00000601958,0.0001447699,0.9942784,0.00002883427,0.001607492,0.0002014003,0.0005033194],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01887375,0.00001555722,0.976674,0.00009498331,0.0006021333,0.0009561107,8.431128e-7,0.001194431,0.001588151],"genre_scores_gemma":[0.9839936,0.0001172605,0.007996733,0.00005368576,0.0001605599,0.000009318731,0.00005797113,0.00006105393,0.007549769],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9686773,"threshold_uncertainty_score":0.9998023,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09279756993421001,"score_gpt":0.2157709403239721,"score_spread":0.1229733703897621,"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."}}