{"id":"https://openalex.org/W3132797628","doi":"https://doi.org/10.48550/arxiv.2102.12425","title":"Synthetic Returns for Long-Term Credit Assignment","display_name":"Synthetic Returns for Long-Term Credit Assignment","publication_year":2021,"publication_date":"2021-02-24","ids":{"openalex":"https://openalex.org/W3132797628","doi":"https://doi.org/10.48550/arxiv.2102.12425","mag":"3132797628"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2102.12425","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2102.12425","pdf_url":"https://arxiv.org/pdf/2102.12425","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2102.12425","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5103096285","display_name":"David Chillo\u0301n Raposo","orcid":"https://orcid.org/0000-0001-8202-5532","observed_orcids":["https://orcid.org/0000-0001-8202-5532"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Raposo, David","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5049000381","display_name":"Samuel Ritter","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ritter, Sam","raw_affiliation_strings":["Google,,,,,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google,,,,,","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112705664","display_name":"Adam Santoro","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Santoro, Adam","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013028446","display_name":"Greg Wayne","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wayne, Greg","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005085749","display_name":"Th\u00e9ophane Weber","orcid":"https://orcid.org/0000-0001-6230-952X","observed_orcids":["https://orcid.org/0000-0001-6230-952X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Weber, Theophane","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083771180","display_name":"Matthew Botvinick","orcid":"https://orcid.org/0000-0001-7758-6896","observed_orcids":["https://orcid.org/0000-0001-7758-6896","https://orcid.org/0009-0000-1902-4000"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Botvinick, Matt","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033135596","display_name":"Hado P. van Hasselt","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"van Hasselt, Hado","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Song, Francis","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Francis","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":31,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11574","display_name":"Artificial Intelligence in Games","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12101","display_name":"Advanced Bandit Algorithms Research","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/temporal-difference-learning","display_name":"temporal difference learning","score":0.9649999737739563},{"id":"https://openalex.org/keywords/delayed-rewards","display_name":"delayed rewards","score":0.6520000100135803},{"id":"https://openalex.org/keywords/impala","display_name":"impala","score":0.6050000190734863},{"id":"https://openalex.org/keywords/interpretability","display_name":"interpretability","score":0.5419999957084656},{"id":"https://openalex.org/keywords/deep-reinforcement-learning","display_name":"deep reinforcement learning","score":0.6549999713897705}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.8510158658027649},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7013661861419678},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.5840951204299927},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5366619825363159},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.4902530312538147},{"id":"https://openalex.org/C188116033","wikidata":"https://www.wikidata.org/wiki/Q2664563","display_name":"Q-learning","level":3,"score":0.4871310889720917},{"id":"https://openalex.org/C196340769","wikidata":"https://www.wikidata.org/wiki/Q7698910","display_name":"Temporal difference learning","level":3,"score":0.4473525583744049},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.43510860204696655},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.42127326130867004},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4135032892227173},{"id":"https://openalex.org/C159423971","wikidata":"https://www.wikidata.org/wiki/Q177251","display_name":"Associative property","level":2,"score":0.4120432734489441},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.3367692828178406},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.18170905113220215},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.15306255221366882},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.1127467155456543},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.07954689860343933},{"id":"https://openalex.org/C121955636","wikidata":"https://www.wikidata.org/wiki/Q4116214","display_name":"Accounting","level":1,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C202444582","wikidata":"https://www.wikidata.org/wiki/Q837863","display_name":"Pure mathematics","level":1,"score":0.0},{"id":"https://openalex.org/C146978453","wikidata":"https://www.wikidata.org/wiki/Q3798668","display_name":"Aerospace engineering","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2102.12425","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2102.12425","pdf_url":"https://arxiv.org/pdf/2102.12425","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2102.12425","doi":"https://doi.org/10.48550/arxiv.2102.12425","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2102.12425","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2102.12425","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2102.12425","pdf_url":"https://arxiv.org/pdf/2102.12425","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320317153","display_name":"DeepMind","ror":"https://ror.org/00971b260"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W3132797628.pdf","grobid_xml":"https://content.openalex.org/works/W3132797628.grobid-xml"},"referenced_works_count":26,"referenced_works":["https://openalex.org/W2064675550","https://openalex.org/W2117726420","https://openalex.org/W2121863487","https://openalex.org/W2145339207","https://openalex.org/W2260756217","https://openalex.org/W2341171179","https://openalex.org/W2402144811","https://openalex.org/W2418628973","https://openalex.org/W2786036274","https://openalex.org/W2786928559","https://openalex.org/W2794752562","https://openalex.org/W2804508831","https://openalex.org/W2809013025","https://openalex.org/W2897023205","https://openalex.org/W2900677074","https://openalex.org/W2908064123","https://openalex.org/W2947137205","https://openalex.org/W2964291307","https://openalex.org/W2970567787","https://openalex.org/W2970811133","https://openalex.org/W2990376820","https://openalex.org/W2991993771","https://openalex.org/W2996695841","https://openalex.org/W3006178546","https://openalex.org/W3013618273","https://openalex.org/W3099551617"],"related_works":["https://openalex.org/W7027420228","https://openalex.org/W4389363365","https://openalex.org/W7105992850","https://openalex.org/W7202159264","https://openalex.org/W7080979222","https://openalex.org/W3094125909","https://openalex.org/W7138444981","https://openalex.org/W4396813826","https://openalex.org/W4394590696","https://openalex.org/W7130728993"],"abstract_inverted_index":{"Since":[0],"the":[1,7,54,70,80,86,99,124,144,192],"earliest":[2],"days":[3],"of":[4,82,126,134],"reinforcement":[5],"learning,":[6,21,52],"workhorse":[8],"method":[9],"for":[10,151],"assigning":[11],"credit":[12,24,67],"to":[13,45,78,85,98,113,184],"actions":[14,33,115,157],"over":[15],"time":[16],"has":[17],"been":[18],"temporal-difference":[19],"(TD)":[20],"which":[22,137],"propagates":[23,66],"backward":[25],"timestep-by-timestep.":[26],"This":[27],"approach":[28],"suffers":[29],"when":[30,39],"delays":[31],"between":[32,58,69],"and":[34,38,60],"rewards":[35],"are":[36,147,158],"long":[37],"intervening":[40],"unrelated":[41],"events":[42],"contribute":[43],"variance":[44],"long-term":[46],"returns.":[47],"We":[48,122,141],"propose":[49],"state-associative":[50],"(SA)":[51],"where":[53],"agent":[55,167],"learns":[56],"associations":[57],"states":[59,84,152],"arbitrarily":[61],"distant":[62],"future":[63],"rewards,":[64],"then":[65,110],"directly":[68],"two.":[71],"In":[72],"this":[73,90],"work,":[74],"we":[75,92,104,161],"use":[76],"SA-learning":[77],"model":[79,91],"contribution":[81,97],"past":[83],"current":[87],"reward.":[88],"With":[89],"can":[93,109],"predict":[94],"each":[95],"state's":[96],"far":[100],"future,":[101],"a":[102,132,172,175,181],"quantity":[103],"call":[105],"\"synthetic":[106],"returns\".":[107],"TD-learning":[108,138],"be":[111],"applied":[112],"select":[114],"that":[116,143,153,163,179],"maximize":[117],"these":[118],"synthetic":[119],"returns":[120],"(SRs).":[121],"demonstrate":[123],"effectiveness":[125],"augmenting":[127],"agents":[128,186],"with":[129,174],"SRs":[130,146],"across":[131],"range":[133],"tasks":[135],"on":[136],"alone":[139],"fails.":[140],"show":[142,162],"learned":[145],"interpretable:":[148],"they":[149],"spike":[150],"occur":[154],"after":[155],"critical":[156],"taken.":[159],"Finally,":[160],"our":[164],"IMPALA-based":[165],"SR":[166],"solves":[168],"Atari":[169],"Skiing":[170],"--":[171,187],"game":[173],"lengthy":[176],"reward":[177],"delay":[178],"posed":[180],"major":[182],"hurdle":[183],"deep-RL":[185],"25":[188],"times":[189],"faster":[190],"than":[191],"published":[193],"state-of-the-art.":[194]},"counts_by_year":[{"year":2026,"cited_by_count":5},{"year":2025,"cited_by_count":3},{"year":2024,"cited_by_count":7},{"year":2023,"cited_by_count":7},{"year":2022,"cited_by_count":4},{"year":2021,"cited_by_count":5}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2021-03-01T00:00:00"}