{"id":"https://openalex.org/W4360764658","doi":"https://doi.org/10.1109/icmla55696.2022.00011","title":"Attention-based Partial Decoupling of Policy and Value for Generalization in Reinforcement Learning","display_name":"Attention-based Partial Decoupling of Policy and Value for Generalization in Reinforcement Learning","publication_year":2022,"publication_date":"2022-12-01","ids":{"openalex":"https://openalex.org/W4360764658","doi":"https://doi.org/10.1109/icmla55696.2022.00011"},"language":"en","primary_location":{"id":"doi:10.1109/icmla55696.2022.00011","doi":"https://doi.org/10.1109/icmla55696.2022.00011","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/icmla55696.2022.00011","pdf_url":null,"source":{"id":"https://openalex.org/S4306419645","display_name":"International Conference on Machine Learning and Applications","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","is_traditional":true,"indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5004876535","display_name":"Nasik Muhammad Nafi","orcid":"https://orcid.org/0000-0002-8492-9217","observed_orcids":["https://orcid.org/0000-0002-8492-9217"]},"institutions":[{"id":"https://openalex.org/I189590672","display_name":"Kansas State University","ror":"https://ror.org/05p1j8758","country_code":"US","type":"education","lineage":["https://openalex.org/I189590672"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Nasik Muhammad Nafi","raw_affiliation_strings":["Kansas State University,Department of Computer Science,Manhattan,KS,USA","Department of Computer Science, Kansas State University, Manhattan, KS, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Kansas State University,Department of Computer Science,Manhattan,KS,USA","institution_ids":["https://openalex.org/I189590672"]},{"raw_affiliation_string":"Department of Computer Science, Kansas State University, Manhattan, KS, USA","institution_ids":["https://openalex.org/I189590672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024820786","display_name":"Creighton Glasscock","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I189590672","display_name":"Kansas State University","ror":"https://ror.org/05p1j8758","country_code":"US","type":"education","lineage":["https://openalex.org/I189590672"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Creighton Glasscock","raw_affiliation_strings":["Kansas State University,Department of Computer Science,Manhattan,KS,USA","Department of Computer Science, Kansas State University, Manhattan, KS, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Kansas State University,Department of Computer Science,Manhattan,KS,USA","institution_ids":["https://openalex.org/I189590672"]},{"raw_affiliation_string":"Department of Computer Science, Kansas State University, Manhattan, KS, USA","institution_ids":["https://openalex.org/I189590672"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5016291944","display_name":"William W.Y. Hsu","orcid":"https://orcid.org/0000-0002-5168-070X","observed_orcids":["https://orcid.org/0000-0002-5168-070X"]},"institutions":[{"id":"https://openalex.org/I189590672","display_name":"Kansas State University","ror":"https://ror.org/05p1j8758","country_code":"US","type":"education","lineage":["https://openalex.org/I189590672"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"William Hsu","raw_affiliation_strings":["Kansas State University,Department of Computer Science,Manhattan,KS,USA","Department of Computer Science, Kansas State University, Manhattan, KS, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Kansas State University,Department of Computer Science,Manhattan,KS,USA","institution_ids":["https://openalex.org/I189590672"]},{"raw_affiliation_string":"Department of Computer Science, Kansas State University, Manhattan, KS, USA","institution_ids":["https://openalex.org/I189590672"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I189590672"],"apc_list":null,"apc_paid":null,"fwci":0.4788,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.65210821,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"15","last_page":"22"},"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","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/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"}}],"keywords":[{"id":"https://openalex.org/keywords/attention-mechanism","display_name":"attention mechanism","score":0.7829999923706055},{"id":"https://openalex.org/keywords/generalization","display_name":"generalization","score":0.6439999938011169},{"id":"https://openalex.org/keywords/overfitting","display_name":"overfitting","score":0.7870000004768372},{"id":"https://openalex.org/keywords/actor-critic-architecture","display_name":"actor-critic architecture","score":0.7570000290870667},{"id":"https://openalex.org/keywords/representation-learning","display_name":"representation learning","score":0.5630000233650208},{"id":"https://openalex.org/keywords/proximal-policy-optimization","display_name":"proximal policy optimization","score":0.7820000052452087}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.845907986164093},{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.8125806450843811},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8084901571273804},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.6597847938537598},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.637820303440094},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4623045027256012},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.440030038356781},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.43531137704849243},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3640696406364441},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.08403438329696655},{"id":"https://openalex.org/C199539241","wikidata":"https://www.wikidata.org/wiki/Q7748","display_name":"Law","level":1,"score":0.07928967475891113},{"id":"https://openalex.org/C17744445","wikidata":"https://www.wikidata.org/wiki/Q36442","display_name":"Political science","level":0,"score":0.0},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0},{"id":"https://openalex.org/C94625758","wikidata":"https://www.wikidata.org/wiki/Q7163","display_name":"Politics","level":2,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icmla55696.2022.00011","doi":"https://doi.org/10.1109/icmla55696.2022.00011","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/icmla55696.2022.00011","pdf_url":null,"source":{"id":"https://openalex.org/S4306419645","display_name":"International Conference on Machine Learning and Applications","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":27,"referenced_works":["https://openalex.org/W2145339207","https://openalex.org/W2260756217","https://openalex.org/W2593658277","https://openalex.org/W2736601468","https://openalex.org/W2752782242","https://openalex.org/W2781726626","https://openalex.org/W2786036274","https://openalex.org/W2894976951","https://openalex.org/W2903181768","https://openalex.org/W2925486737","https://openalex.org/W2937843571","https://openalex.org/W2949718784","https://openalex.org/W2962858109","https://openalex.org/W2966537656","https://openalex.org/W2969262604","https://openalex.org/W2970214542","https://openalex.org/W2994073215","https://openalex.org/W2996110979","https://openalex.org/W3012502279","https://openalex.org/W3039141380","https://openalex.org/W3092744199","https://openalex.org/W3103999165","https://openalex.org/W3118409705","https://openalex.org/W3129741749","https://openalex.org/W3166388069","https://openalex.org/W3169036355","https://openalex.org/W6741002519"],"related_works":["https://openalex.org/W3129741749","https://openalex.org/W7124315350","https://openalex.org/W7133186840","https://openalex.org/W4312387554","https://openalex.org/W4385488677","https://openalex.org/W7197621336","https://openalex.org/W7131379121","https://openalex.org/W7117550058","https://openalex.org/W3166968915","https://openalex.org/W4385438429"],"abstract_inverted_index":{"In":[0],"this":[1],"work,":[2],"we":[3],"introduce":[4],"Attention-based":[5],"Partially":[6],"Decoupled":[7],"Actor-Critic":[8],"(APDAC),":[9],"an":[10,136,165],"actor-critic":[11,33],"architecture":[12,98,107],"for":[13,73,130],"generalization":[14,64,85,181],"in":[15,95],"reinforcement":[16],"learning,":[17],"which":[18],"partially":[19],"separates":[20],"the":[21,24,41,67,84,89,111,123,127,149,155,158,173,178,191],"policy":[22,42,74,131,150],"and":[23,43,52,75,82,100,125,132,151,184,201],"value":[25,44,76,152],"functions.":[26,45],"To":[27],"learn":[28],"directly":[29],"from":[30],"images,":[31],"traditional":[32],"architectures":[34,116],"use":[35],"a":[36,47,105],"shared":[37,48],"network":[38,124],"to":[39,59,79,139],"represent":[40],"While":[46],"representation":[49,142],"allows":[50],"parameter":[51],"feature":[53],"sharing,":[54],"it":[55],"can":[56,77],"also":[57],"lead":[58],"overfitting":[60,81],"that":[61,108,153,169,187],"catastrophically":[62],"damages":[63],"performance.":[65],"On":[66],"other":[68],"hand,":[69],"two":[70],"separate":[71],"networks":[72],"help":[78],"avoid":[80],"reduce":[83],"gap,":[86],"but":[87],"at":[88,210],"cost":[90],"of":[91,97,114,122,148,157],"added":[92],"complexity":[93],"both":[94,115],"terms":[96],"design":[99],"computation":[101],"time.":[102,205],"APDAC":[103,134,170],"is":[104,188,208],"hybrid":[106],"builds":[109],"upon":[110],"combined":[112],"strengths":[113],"by":[117],"sharing":[118],"initial":[119],"layer":[120],"blocks":[121],"separating":[126],"later":[128],"ones":[129],"value.":[133],"incorporates":[135],"attention":[137],"mechanism":[138],"enable":[140],"robust":[141],"learning.":[143],"We":[144],"present":[145],"meaningful":[146],"visualization":[147],"explains":[154],"perception":[156],"trained":[159],"agent.":[160],"Our":[161,206],"empirical":[162],"analysis,":[163],"including":[164],"ablation":[166],"study,":[167],"shows":[168],"significantly":[171],"outperforms":[172],"standard":[174],"PPO":[175],"baseline":[176],"on":[177],"challenging":[179],"RL":[180],"benchmark":[182],"Procgen":[183],"achieves":[185],"performance":[186],"competitive":[189],"with":[190],"recent":[192],"state-of-the-art":[193],"method":[194],"(IDAAC)":[195],"while":[196],"using":[197],"fewer":[198],"convolutional":[199],"layers":[200],"requiring":[202],"less":[203],"computational":[204],"code":[207],"available":[209],"https://github.com/nasiknafi/apdac.":[211]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":2}],"updated_date":"2026-10-26T11:51:55.731654","created_date":"2023-03-25T00:00:00"}