{"id":"https://openalex.org/W4414359584","doi":"https://doi.org/10.24963/ijcai.2025/609","title":"Learning to Explain: Towards Human-Aligned Explainability in Deep Reinforcement Learning via Attention Guidance","display_name":"Learning to Explain: Towards Human-Aligned Explainability in Deep Reinforcement Learning via Attention Guidance","publication_year":2025,"publication_date":"2025-09-01","ids":{"openalex":"https://openalex.org/W4414359584","doi":"https://doi.org/10.24963/ijcai.2025/609"},"language":"en","primary_location":{"id":"doi:10.24963/ijcai.2025/609","is_oa":false,"landing_page_url":"https://doi.org/10.24963/ijcai.2025/609","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"},"type":"article","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/A5109640208","display_name":"Bokai Ji","orcid":"https://orcid.org/0009-0008-4676-1629"},"institutions":[{"id":"https://openalex.org/I149594827","display_name":"Xidian University","ror":"https://ror.org/05s92vm98","country_code":"CN","type":"education","lineage":["https://openalex.org/I149594827"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Bokai Ji","raw_affiliation_strings":["Xidian University"],"affiliations":[{"raw_affiliation_string":"Xidian University","institution_ids":["https://openalex.org/I149594827"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5009139939","display_name":"Guangxia Li","orcid":"https://orcid.org/0000-0003-2158-8723"},"institutions":[{"id":"https://openalex.org/I149594827","display_name":"Xidian University","ror":"https://ror.org/05s92vm98","country_code":"CN","type":"education","lineage":["https://openalex.org/I149594827"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Guangxia Li","raw_affiliation_strings":["Xidian University"],"affiliations":[{"raw_affiliation_string":"Xidian University","institution_ids":["https://openalex.org/I149594827"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043356063","display_name":"Yulong Shen","orcid":"https://orcid.org/0000-0002-8448-705X"},"institutions":[{"id":"https://openalex.org/I149594827","display_name":"Xidian University","ror":"https://ror.org/05s92vm98","country_code":"CN","type":"education","lineage":["https://openalex.org/I149594827"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yulong Shen","raw_affiliation_strings":["Xidian University"],"affiliations":[{"raw_affiliation_string":"Xidian University","institution_ids":["https://openalex.org/I149594827"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101487120","display_name":"Gang Xiao","orcid":"https://orcid.org/0000-0002-9259-6790"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gang Xiao","raw_affiliation_strings":["National Key Laboratory for Complex Systems Simulation, China"],"affiliations":[{"raw_affiliation_string":"National Key Laboratory for Complex Systems Simulation, China","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5109640208"],"corresponding_institution_ids":["https://openalex.org/I149594827"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.13951673,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"5472","last_page":"5479"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.9143000245094299,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.9143000245094299,"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/interpretability","display_name":"Interpretability","score":0.8422999978065491},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.6960999965667725},{"id":"https://openalex.org/keywords/cognition","display_name":"Cognition","score":0.5322999954223633},{"id":"https://openalex.org/keywords/relevance","display_name":"Relevance (law)","score":0.4862000048160553},{"id":"https://openalex.org/keywords/component","display_name":"Component (thermodynamics)","score":0.4120999872684479},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.3499999940395355}],"concepts":[{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.8422999978065491},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6960999965667725},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.628000020980835},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5848000049591064},{"id":"https://openalex.org/C169900460","wikidata":"https://www.wikidata.org/wiki/Q2200417","display_name":"Cognition","level":2,"score":0.5322999954223633},{"id":"https://openalex.org/C158154518","wikidata":"https://www.wikidata.org/wiki/Q7310970","display_name":"Relevance (law)","level":2,"score":0.4862000048160553},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.4120999872684479},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3499999940395355},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3449999988079071},{"id":"https://openalex.org/C2778732403","wikidata":"https://www.wikidata.org/wiki/Q815577","display_name":"Ignorance","level":2,"score":0.33390000462532043},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.31929999589920044},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.2808000147342682},{"id":"https://openalex.org/C67203356","wikidata":"https://www.wikidata.org/wiki/Q1321905","display_name":"Reinforcement","level":2,"score":0.2687000036239624},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.2646999955177307},{"id":"https://openalex.org/C188147891","wikidata":"https://www.wikidata.org/wiki/Q147638","display_name":"Cognitive science","level":1,"score":0.26339998841285706},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.25999999046325684},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.2540000081062317}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.24963/ijcai.2025/609","is_oa":false,"landing_page_url":"https://doi.org/10.24963/ijcai.2025/609","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0],"advances":[1],"in":[2,75,146],"explainable":[3,153],"deep":[4],"reinforcement":[5],"learning":[6],"(DRL)":[7],"have":[8],"provided":[9],"insights":[10],"into":[11,65],"the":[12,25,44,49,66,94,98,127],"reasoning":[13],"behind":[14],"decisions":[15],"made":[16],"by":[17,69,91,121],"DRL":[18,71,95,154],"agents.":[19],"However,":[20],"existing":[21,152],"methods":[22],"often":[23],"overlook":[24],"subjective":[26],"nature":[27],"of":[28,48],"explanations":[29,51,90,102,149],"and":[30,37,46,97,103,140],"fail":[31],"to":[32,42,87,151],"consider":[33],"human":[34,54,63,104],"cognitive":[35],"styles":[36],"preferences.":[38],"Such":[39],"ignorance":[40],"tends":[41],"reduce":[43],"interpretability":[45],"relevance":[47],"generated":[50,101],"from":[52],"a":[53,76,110],"evaluator's":[55],"perspective.":[56],"To":[57],"address":[58],"this":[59],"issue,":[60],"we":[61],"introduce":[62],"cognition":[64],"explaining":[67,119],"procedure":[68],"integrating":[70],"with":[72],"attention":[73],"guidance":[74],"novel":[77],"manner.":[78],"The":[79],"proposed":[80],"concept":[81,114],"proximal":[82],"policy":[83],"optimization":[84],"(Concept-PPO)":[85],"learns":[86],"generate":[88],"human-aligned":[89,148],"jointly":[92],"optimizing":[93],"performance":[96],"discrepancy":[99],"between":[100],"annotations.":[105],"Its":[106],"key":[107],"component":[108],"is":[109],"specially":[111],"designed":[112],"spatial":[113],"transformer":[115],"that":[116,131],"can":[117],"enhance":[118],"efficiency":[120],"premasking":[122],"decision-irrelevant":[123],"information.":[124],"Experiments":[125],"on":[126],"ATARI":[128],"benchmark":[129],"demonstrate":[130],"Concept-PPO":[132],"achieves":[133],"better":[134],"policies":[135],"than":[136],"its":[137,144],"black-box":[138],"counterparts,":[139],"user":[141],"studies":[142],"confirm":[143],"superiority":[145],"generating":[147],"compared":[150],"methods.":[155]},"counts_by_year":[],"updated_date":"2026-03-07T16:01:11.037858","created_date":"2025-10-10T00:00:00"}